Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction [First ed.] 9780851994635, 9780851997124, 0851994636


236 112 3MB

English Pages 474 [489] Year 2000

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Contents......Page 6
Contributors......Page 10
Foreword......Page 14
I Regulation of Feed Intake......Page 16
1 Integration of Learning and Metabolic Signals into a Theory of Dietary Choice and Food Intake......Page 18
2 Mathematical Models of Food Intake and Metabolism in Ruminants......Page 36
3 Control of Salivation and Motility of the Reticulorumen by the Brain in Sheep......Page 56
II Rumen Microbiology and Fermentation......Page 74
4 Molecular Ecology and Diversity in Gut Microbial Ecosystems......Page 76
5 Microbial Adherence to the Plant Cell Wall and Enzymatic Hydrolysis......Page 94
6 The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism......Page 114
III Nutrient Absorption and Splanchnic Metabolism......Page 130
7 Tissue, Cellular and Molecular Aspects of Peptide Absorption and Utilization......Page 132
8 Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver......Page 146
9 The Liver: Integrator of Nitrogen Metabolism......Page 164
IV Tissue Maintenance and Utilization of Endogenous Body Reserves......Page 184
10 Adipose Tissue: Beyond an Energy Reserve......Page 186
11 Regulation of Growth and Metabolism During Postnatal Development......Page 202
12 Direct Effects of Photoperiod on Lipid Metabolism, Leptin Synthesis and Milk Secretion in Adult Sheep......Page 220
V Tissue Growth......Page 240
13 Muscle Growth and Genetic Regulation......Page 242
14 Control and Manipulation of Hyperplasia and Hypertrophy in Muscle Tissue......Page 252
15 Regulation of Protein Synthesis for Wool Growth......Page 270
VI Reproduction, Pregnancy and Lactation......Page 288
16 Regulation of Macronutrient Partitioning between Maternal and Conceptus Tissues in the Pregnant Ruminant......Page 290
17 The Thermal Physiology of the Ruminant Fetus......Page 310
18 Regulation of Nutrient Partitioning During Lactation: Homeostasis and Homeorhesis Revisited......Page 326
19 The Insulin-like Growth Factor (IGF) System in the Mammary Gland: Role of IGFBP-3 Binding Protein......Page 344
20 Integrating the Effects of Genotype and Nutrition on Utilization of Body Reserves During Lactation of Dairy Cattle......Page 368
VII Ruminant Physiology and Genetics......Page 386
21 Genetic Manipulation of Ruminant Biochemistry and Physiology for Improved Productivity: Current Status and Future Potential......Page 388
22 Genetics of Rumen Microorganisms: Gene Transfer, Genetic Analysis and Strain Manipulation......Page 404
23 Nutrient–Gene Interactions: Future Potential and Applications......Page 424
VIII Host Resistance to Parasites and Pathogens......Page 438
24 Host Resistance to Gastrointestinal Parasites of Sheep......Page 440
25 Host Resistance to Fleece Rot and Blowfly Strike......Page 452
26 Host Resistance to Mastitis......Page 464
Index......Page 478
Recommend Papers

Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction [First ed.]
 9780851994635, 9780851997124, 0851994636

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

RUMINANT PHYSIOLOGY: Digestion, Metabolism, Growth and Reproduction

Edited by P.B. Cronjé

CABI Publishing

RUMINANT PHYSIOLOGY Digestion, Metabolism, Growth and Reproduction

Dedication This volume is dedicated to the memory of the late Dr F.M.C. Gilchrist.

RUMINANT PHYSIOLOGY Digestion, Metabolism, Growth and Reproduction

Edited by

P.B. Cronjé Department of Animal and Wildlife Sciences University of Pretoria Pretoria South Africa Associate Editors

E.A. Boomker P.H. Henning W. Schultheiss J.G. van der Walt

CABI Publishing

CABI Publishing is a division of CAB International CABI Publishing CAB International Wallingford Oxon OX10 8DE UK Tel: +44 (0)1491 832111 Fax: +44 (0)1491 833508 Email: [email protected] Web site: http://www.cabi.org

CABI Publishing 10 E 40th Street Suite 3203 New York, NY 10016 USA Tel: +1 (212) 481 7018 Fax: +1 (212) 686 7993 Email: [email protected]

© CAB International 2000. All rights reserved. No part of this publication may be reproduced in any form or by any means, electronically, mechanically, by photocopying, recording or otherwise, without the prior permission of the copyright owners. A catalogue record for this book is available from the British Library, London, UK Library of Congress Cataloging-in-Publication Data Ruminant physiology : digestion, metabolism, growth, and reproduction / edited by P. Cronje ; assoc. editors, E.A. Boomker … [et al.]. p. cm. Includes bibliographical references and index. ISBN 0-85199-463-6 (alk. paper) 1. Ruminants--Physiology--Congresses. I. Cronjé, P. (Pierre) II. Boomker, E. A. QL737.U5 R868 2000 571.1¢ 963- - dc21

ISBN 0 85199 463 6

Typeset in 10/12pt Garamond by Columns Design Ltd, Reading Printed and bound in the UK by Biddles Ltd, Guildford and King’s Lynn

00–023661

Contents

Contributors Foreword Part I Regulation of Feed Intake 1 Integration of Learning and Metabolic Signals into a Theory of Dietary Choice and Food Intake J.M. Forbes and F.D. Provenza 2 Mathematical Models of Food Intake and Metabolism in Ruminants A.W. Illius, N.S. Jessop and M. Gill 3 Control of Salivation and Motility of the Reticulorumen by the Brain in Sheep W.L. Grovum and J.S. Gonzalez Part II Rumen Microbiology and Fermentation

ix xiii 1 3 21

41 59

4 Molecular Ecology and Diversity in Gut Microbial Ecosystems R.I. Mackie, R.I. Aminov, B.A. White and C.S. McSweeney

61

5 Microbial Adherence to the Plant Cell Wall and Enzymatic Hydrolysis C.W. Forsberg, E. Forano and A. Chesson

79

6 The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism M. Morrison

99

Part III Nutrient Absorption and Splanchnic Metabolism 7 Tissue, Cellular and Molecular Aspects of Peptide Absorption and Utilization K.E. Webb, Jr

115 117 v

Contents

vi

8 Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver C.J. Seal and D.S. Parker

131

9 The Liver: Integrator of Nitrogen Metabolism G.E. Lobley, G.D. Milano and J.G. van der Walt

149

Part IV Tissue Maintenance and Utilization of Endogenous Body Reserves

169

10 Adipose Tissue: Beyond an Energy Reserve R.G. Vernon and K.L. Houseknecht

171

11 Regulation of Growth and Metabolism During Postnatal Development B.H. Breier, M.H. Oliver and B.W. Gallaher

187

12 Direct Effects of Photoperiod on Lipid Metabolism, Leptin Synthesis and Milk Secretion in Adult Sheep Y. Chilliard and F. Bocquier

205

Part V Tissue Growth

225

13 Muscle Growth and Genetic Regulation J.J. Bass, M. Sharma, J. Oldham and R. Kambadur

227

14 Control and Manipulation of Hyperplasia and Hypertrophy in Muscle Tissue P.J. Buttery, J.M. Brameld and J.M. Dawson

237

15 Regulation of Protein Synthesis for Wool Growth N.R. Adams, S. Liu and D.G. Masters

255

Part VI Reproduction, Pregnancy and Lactation

273

16 Regulation of Macronutrient Partitioning between Maternal and Conceptus Tissues in the Pregnant Ruminant A.W. Bell and R.A. Ehrhardt

275

17 The Thermal Physiology of the Ruminant Fetus H. Laburn, A. Faurie and D. Mitchell

295

18 Regulation of Nutrient Partitioning During Lactation: Homeostasis and Homeorhesis Revisited D.E. Bauman

311

19 The Insulin-like Growth Factor (IGF) System in the Mammary Gland: Role for IGFBP-3 Binding Protein C.R. Baumrucker

329

20 Integrating the Effects of Genotype and Nutrition on Utilization of Body Reserves During Lactation of Dairy Cattle J.P. McNamara

353

Contents

vii

Part VII Ruminant Physiology and Genetics

371

21 Genetic Manipulation of Ruminant Biochemistry and Physiology for Improved Productivity: Current Status and Future Potential K.A. Ward

373

22 Genetics of Rumen Microorganisms: Gene Transfer, Genetic Analysis and Strain Manipulation H.J. Flint and K.P. Scott

389

23 Nutrient–Gene Interactions: Future Potential and Applications P.B. Cronjé

409

Part VIII Host Resistance to Parasites and Pathogens

423

24 Host Resistance to Gastrointestinal Parasites of Sheep S.J. McClure, D.L. Emery and J.W. Steel

425

25 Host Resistance to Fleece Rot and Blowfly Strike I.G. Colditz and R.L. Tellam

437

26 Host Resistance to Mastitis K. Persson Waller

449

Index

463

Contributors

N.R. Adams, CSIRO Division of Animal Production and CRC for Premium Quality Wool, Wembley, 6014 Western Australia R.I. Aminov, Department of Animal Sciences and Division of Nutritional Sciences, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA J.J. Bass, AgResearch, Ruakura Agricultural Research Centre, Private Bag 3123, Hamilton, New Zealand D.E. Bauman, Department of Animal Science, Cornell University, Ithaca, NY 14853, USA C.R. Baumrucker, Department of Dairy and Animal Science, Penn State University, 302 Henning Building, University Park, PA 16802, USA A.W. Bell, Department of Animal Science, Cornell University, Ithaca, NY 14853, USA F. Bocquier, Adipose Tissue and Milk Lipids Team, Herbivore Research Unit, INRATheix, 63122 St Genès Champanelle, France J.M. Brameld, Division of Nutritional Biochemistry, School of Biological Sciences, University of Nottingham, Sutton Bonington Campus, Loughborough, Leicestershire LEI2 5RD, UK B.H. Breier, Research Centre for Developmental Medicine and Biology, Faculty of Medicine and Health Science, University of Auckland, Private Bag 92019, Auckland, New Zealand P.J. Buttery, Division of Nutritional Biochemistry, School of Biological Sciences, University of Nottingham, Sutton Bonington Campus, Loughborough, Leicestershire LE12 5RD, UK A. Chesson, Rowett Research Institute, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, UK Y. Chilliard, Adipose Tissue and Milk Lipids Team, Herbivore Research Unit, INRATheix, 63122 St Genès Champanelle, France I.G. Colditz, CSIRO Animal Production, Pastoral Research Laboratory, Armidale, NSW 2350, Australia

ix

x

Contributors

P.B. Cronjé, Department of Animal and Wildlife Sciences, University of Pretoria, Pretoria 0002, South Africa J.M. Dawson, Division of Nutritional Biochemistry, School of Biological Sciences, University of Nottingham, Sutton Bonington Campus, Loughborough, Leicestershire LE12 5RD, UK R.A. Ehrhardt, Department of Animal Science, Cornell University, Ithaca, NY 14853, USA D.L. Emery, CSIRO Animal Production, McMaster Laboratory, Prospect, NSW 2148, Australia A. Faurie, Department of Physiology and Brain Function Research Unit, University of the Witwatersrand, 7 York Road, Parktown, Johannesburg 2193, South Africa H.J. Flint, Rowett Research Institute, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, UK E. Forano, Laboratoire de Microbiologie, INRA CR de Clermont-Ferrand – Theix, 63122 St Genès Champanelle, France J.M. Forbes, Centre for Animal Sciences, Leeds Institute of Biotechnology and Agriculture, University of Leeds, Leeds LS2 9JT, UK C.W. Forsberg, Department of Microbiology, University of Guelph, Guelph, Ontario N1G 2W1, Canada B.W. Gallaher, Research Centre for Developmental Medicine and Biology, Faculty of Medicine and Health Science, University of Auckland, Private Bag 92019, Auckland, New Zealand M. Gill, NR International, Central Avenue, Chatham Maritime, Kent ME4 4TB, UK J.S. Gonzalez, Department of Animal Production, University of Leon, 24071 Leon, Spain W.L. Grovum, Department of Biomedical Sciences, Ontario Veterinary College, University of Guelph, Guelph, Ontario, N1G 2W1, Canada K.L. Houseknecht, Animal Health Drug Discovery, Pfizer Inc., Groton, CT 063408002, USA A.W. Illius, Division of Biological Sciences, University of Edinburgh, West Mains Road, Edinburgh EH9 3JT, UK N.S. Jessop, Division of Biological Sciences, University of Edinburgh, West Mains Road, Edinburgh EH9 3JT, UK R. Kambadur, AgResearch, Ruakura Agricultural Research Centre, Private Bag 3123, Hamilton, New Zealand H. Laburn, Department of Physiology and Brain Function Research Unit, University of the Witwatersrand, 7 York Road, Parktown, Johannesburg 2193, South Africa S. Liu, CSIRO Division of Animal Production and CRC for Premium Quality Wool, Wembley, 6014 Western Australia G.E. Lobley, Rowett Research Institute, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, UK R.I. Mackie, Department of Animal Sciences and Division of Nutritional Sciences, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA D.G. Masters, CSIRO Division of Animal Production and CRC for Premium Quality Wool, Wembley, 6014 Western Australia S.J. McClure, CSIRO Animal Production, McMaster Laboratory, Prospect, NSW 2148, Australia J.P. McNamara, Department of Animal Sciences, 233 Clark Hall, Washington State University, PO Box 646351, Pullman, WA 99164-6351, USA

Contributors

xi

C.S. McSweeney, CSIRO Tropical Agriculture, Long Pocket Laboratory, Indooroopilly, Queensland 4068, Australia G.D. Milano, Departamento de Fisiopatologia, Facultad de Ciencias Veterinarias UNCPBA, Campus Iniversitario, Paraje Arroyo Seco (7000) Tandil, Argentina D. Mitchell, Department of Physiology and Brain Function Research Unit, University of the Witwatersrand, 7 York Road, Parktown, Johannesburg 2193, South Africa M. Morrison, Department of Animal Science, School of Biological Sciences and Center for Biotechnology, University of Nebraska, Lincoln, NE 68583-0908, USA J. Oldham, AgResearch, Ruakura Agricultural Research Centre, Private Bag 3123, Hamilton, New Zealand M.H. Oliver, Research Centre for Developmental Medicine and Biology, Faculty of Medicine and Health Science, University of Auckland, Private Bag 92019, Auckland, New Zealand D.S. Parker, Department of Biological and Nutritional Sciences, Faculty of Agriculture and Biological Science, University of Newcastle upon Tyne, Newcastle upon Tyne NE1 7RU, UK: present address; Novius Europe s.a./n.v., Rue Gulledellestraat 94, B-1200 Brussels, Belgium K. Persson Waller, Swedish University of Agricultural Sciences, Faculty of Veterinary Medicine, Department of Obstetrics and Gynaecology, P0 Box 7039 Uppsala, S-750 07 Sweden F.D. Provenza, Department of Rangeland Resources, Utah State University, Logan, Utah 84322–5230, USA K.P. Scott, Rowett Research Institute, Greenburn Road, Bucksburn, Aberdeen AB21 9SB, UK C.J. Seal, Department of Biological and Nutritional Sciences, Faculty of Agriculture and Biological Science, University of Newcastle upon Tyne, Newcastle upon Tyne NE1 7RU, UK M. Sharma, AgResearch, Ruakura Agricultural Research Centre, Private Bag 3123, Hamilton, New Zealand J.W. Steel, CSIRO Animal Production, McMaster Laboratory, Prospect, NSW 2148, Australia R.L. Tellam, CSIRO Tropical Agriculture, Longpocket Laboratory, Indooroopilly, Queensland 4068, Australia R.G. Vernon, Hannah Research Institute, Ayr KA6 5HL, UK J.G. van der Walt, Department of Veterinary Physiology, Faculty of Veterinary Science, Private Bag X04, Onderstepoort 0110, South Africa K.A. Ward, CSIRO Animal Production, LB1, Delivery Centre, Blacktown, NSW 2148 Australia B.A. White, Department of Animal Sciences and Division of Nutritional Sciences, University of Illinois at Urbana-Champaign, 1207 W. Gregory Drive, Urbana, IL 61801, USA K.E. Webb, Jr, Department of Animal and Poultry Sciences, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061-0306, USA

Foreword

The IX International Symposium on Ruminant Physiology was held in Pretoria, South Africa, during October 1999, and followed the traditions set by the previous meetings held in Nottingham (1960), Ames (1965), Cambridge (1969), Sydney (1974), Clermont-Ferrand (1979), Banff (1984), Sendai (1989) and Willingen (1994). The event was attended by 250 delegates from 28 countries. The plenary papers are published in this volume, and the 242 poster communications were published in the South African Journal of Animal Science. The central issue that emerged from this symposium was that new technologies, notably molecular biology and modelling, have become important research tools for the physiologist. It was, however, apparent that more research is needed to relate advances in these technologies and in our understanding of fundamental physiological mechanisms to the solution of practical problems. The papers reiterated that animals, as free-living organisms, have an inherent ability to select for different nutrients and to control nutrient partitioning between different tissues, but that this can be modified and manipulated by human intervention. The important question of whether tissues are in competition for nutrients or whether the partitioning of nutrients is an integral part of coordinating the optimal use of nutrients will, no doubt, elicit much new research. There has been a clear shift from the traditional nutritional input–fermentation approach to rumen microbiology towards a molecular ecology approach, and a new horizon has appeared with regard to our quest to understand host–parasite relationships. Thanks are expressed to the members of the organizing committee, sub-committees and the many willing helpers. The sponsors deserve a special mention: their contributions made it possible to give financial support to deserving delegates, to publish the poster abstracts and to produce this volume. An international guiding committee was constituted to consider the future of the ISRP and the offers to host the X ISRP. The members of the committee are: Dr John Bass, Prof. Alan Bell, Prof. Giuseppi Bertoni, Prof. Peter Buttery, Prof. Norman Casey (convenor), Dr Yves Chilliard, Prof. Pierre Cronjé, Prof. Jong Ha, Dr Jan Hofmeyr

xiv

Foreword

(WAAP Vice-president), Dr Heinz Meissner, Prof. Y Obara and Prof. Wolfgang von Engelhardt. The committee formulated a number of guiding principles for future ISRP meetings. These are: that the meeting should retain the character of previous symposia; that the focus should remain on the physiology of livestock, reviewing advances over the previous 5 years and setting directions for the next period; that comparative physiology and the impact of advances in physiology on products and sensitive consumer issues are important; that the venue for meetings should be situated where there is a core of established ruminant physiologists who could organize the symposium and, in particular, attend to the scientific programme and publish the proceedings; that the symposium should be easily accessible to young scientists and scientifically developing communities. After considering several invitations and taking the principles agreed upon into consideration, the hosting of the X ISRP was awarded to Denmark. Norman H. Casey (Chairman: Organizing Committee of the IX ISRP)

I

Regulation of Feed Intake

1

Integration of Learning and Metabolic Signals into a Theory of Dietary Choice and Food Intake J.M. FORBES1 AND F.D. PROVENZA2 1Centre

for Animal Sciences, Leeds Institute of Biotechnology and Agriculture, University of Leeds, Leeds, UK; 2Department of Rangeland Resources, Utah State University, Logan, Utah, USA

Introduction The challenge of understanding how diet selection and food intake are controlled is one that occupies an important place in the fields of nutrition, physiology and psychology. In the case of ruminant animals there are two special reasons for our interest in the subject: the complexities of the digestive system and consequent metabolic peculiarities; and the agricultural and ecological importance of the sub-order. Despite several decades of intensive study there is still no consensus on how intake is controlled (Fisher, 1996), nor is there agreement about the way in which animals determine which food(s) to eat when a choice is available. The past few years have seen the publication of sufficient new evidence to allow us to advance our hypotheses about the control of food intake and diet selection. Firstly we review advances in our understanding of the role of learning in determining preferences and aversions for foods by ruminants; we then summarize the ways in which the central nervous system (CNS) is informed about digestive and metabolic processes; discuss the day-to-day variation in intake as an enabling factor in the linking of learning with the physiological consequences of eating; and finally propose how learning and metabolic information are brought together to provide testable hypotheses of the control of diet selection and voluntary food intake. We take it as axiomatic that long- and short-term regulation of intake are interwoven and do not attempt to differentiate between the two.

Learned associations between the sensory properties of a food and the metabolic consequences of eating that food This section presents recent evidence to reinforce the concept that ruminant animals learn to associate the post-ingestive consequences of eating a food with the sensory properties of that food and that they use such conditioned preferences and aversions to direct their selection between foods. © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

3

4

J.M. Forbes and F.D. Provenza

Adaptation of choices of foods in order to avoid excessive intakes of toxins and to ensure adequate intakes of essential nutrients In establishing that ruminants can learn to choose between foods to avoid toxicity it is logical to start with an overtly toxic substance, i.e. LiCl, which has been widely used in conditioned aversion studies. Sheep find LiCl, injected or in the food, to be unpleasant (Dutoit et al., 1991) as it induces a conditioned taste aversion, the strength of which is proportional to the dose administered. Feeding neophobia also increases as a function of the LiCl dose associated with the last novel food encountered. When sheep and goats were offered food containing 2% LiCl, their daily intake after the third day fluctuated about a level that resulted in a LiCl dose of 39 mg kg⫺1 for sheep and 27 mg kg⫺1 for goats, i.e. similar to the doses causing mild aversion in rats and human beings. The word ‘toxin’ is usually reserved for a substance that causes obvious signs of discomfort or distress. However, all dietary components are capable of acting as toxins, if present in great excess over requirements. Even a mild excess can generate aversion as toxins do not have to be consciously sensed in order for their effects to be relayed to the CNS and to have the potential to influence learned aversion. Equally, a deficiency of an essential nutrient can form the unconditioned stimulus for the development of food aversions. An example is provided by Hills et al. (1998) in which sheep either replete or depleted in sulphur were offered foods with different contents of sulphur. Replete sheep given high- and low-sulphur foods initially ate at random but within 2 days reduced the proportion of the high-sulphur food to achieve a sulphur concentration in the total diet very close to that thought to be optimal. Conversely, depleted sheep initially ate a high proportion of the high-sulphur food but later reduced the sulphur content chosen until it stabilized at the optimum level. Other examples of non-random diet selection in order to control the intake of a ‘nutrient’ are: protein (Kyriazakis and Oldham, 1993), sodium (Denton, 1982), energy (Burritt and Provenza, 1992) and oxalic acid (Kyriazakis et al., 1998).

Ruminants learn preferences for a food flavour associated with infusions that correct deficiency; the same nutrient given to excess leads to avoidance of the associated flavour In order to demonstrate unequivocally that such appetites are dependent on learned associations between the sensory properties of the foods and their nutritive value it is necessary to divorce the flavour of the food from its yield of nutrients. This can be done by offering animals a distinctive food and at the same time giving a nutrient by a route that bypasses the mouth, usually intraruminal infusion. In one such experiment with lambs (Villalba and Provenza, 1997a) one flavour was paired with rumen infusion of starch (2.5–9.4% of daily digestible energy (DE) intake) and another flavour with control. Subsequent preference was strongly for the starch-paired flavour, even 8 weeks after infusions had stopped. Starch is rapidly fermented to volatile fatty acids, predominantly propionic, in the rumen. Propionate absorption is likely to be insufficient for glucose synthesis in straw-fed animals so the hypothesis was tested that the supply of this limiting nutrient would induce a preference for the flavour of food eaten

Integration of Learning and Metabolic Signals

5

during supplementation (Villalba and Provenza, 1996). Even though the propionate supplied was equivalent to no more than 1.4% of the daily metabolizable energy (ME) intake, after 8 days of conditioning the sheep had developed a strong preference for food flavoured with that flavour given during supplementation. It was shown that the preference was induced by the propionate rather than the sodium or osmolality of the infusions (Villalba and Provenza, 1996; Villalba and Provenza, 1997b). We can conclude that a single nutrient can induce a preference or an aversion to the flavour it was paired with during training, depending on the rate of administration in relation to the animal’s requirements.

Continuum from deficiency, through sufficiency, to excess for each nutrient Several recent experiments have addressed the question: Do ruminants prefer a flavour associated with an intermediate, optimum content of a nutrient over flavours associated with the same nutrient present in excess or deficiency? Arsenos and Kyriazakis (1999) have demonstrated a continuum between conditioned preferences and aversions in sheep to flavoured foods associated with doses of casein from 9 to 53 g given by gavage. The lower two doses led to conditioned flavour preferences, presumably because they alleviated a N deficiency, while the higher two doses led to conditioned flavour aversions, presumably being sensed as toxic overdoses. The authors observe that the existence of a continuum of flavour preferences and aversions created by different amounts of the same nutrient source could be the basis of how ruminants select a diet which meets their nutrient requirement at a particular point in time. Sheep preferred a flavour paired with intraruminal administration of acetate at several doses (4, 8 or 12% of daily DE intake) or propionate (4% of daily DE intake), but became averse to a flavour paired with higher doses of propionate (12% of daily DE intake). This again suggests a role for learning about different concentrations of metabolites in the control of diet selection (Villalba and Provenza, 1997b). There is thus evidence that a food that the animal believes alleviates a deficiency becomes preferred over other foods, while one thought to be excessive in the same nutrient becomes aversive. Under natural conditions such responses would lead to ‘nutritional wisdom’, i.e. eating a mixture of foods which most closely meets the animal’s nutrient requirements.

Ratio in which nutrients are supplied by different foods affects dietary choice There are some situations in which the ratio of nutrients being absorbed from the digestive tract is such as to induce metabolic imbalance. When acetate and propionate were infused together into the rumen of sheep, conditioned preferences were demonstrated for the associated flavoured wheat straw but the preference was greater when the ratio of acetate:propionate in the infusate was 55:45 than when it was 75:25 (Villalba and Provenza, 1997b). It is likely that straw-fed lambs, with a high ratio of acetate:propionate produced by normal ruminal fermentation, would be deficient in glucose and this would be better alleviated by the mixture with the higher proportion of propionate.

6

J.M. Forbes and F.D. Provenza

It has been proposed that certain types of diet provide imbalances between energy and protein supply at different times of day, even though they may be balanced overall. Kyriazakis and Oldham (1997) set out to test whether such asynchrony of nutrient supply would influence diet selection, relative to a food designed to provide the same nutrient supply in a synchronous manner. Foods were formulated to provide rapidlyor slowly-fermentable energy with high or low rumen degradable protein (RDP) – all foods had the same calculated contents of ME and metabolizable protein (MP). When choices were offered, the proportion of the low-RDP food in the selected diet was lower when the carbohydrate source was rapidly – rather than slowly – fermentable, which is consistent with the hypothesis that ruminants learn to select a mixture of foods that minimizes metabolic imbalance. Lambs discriminate between the post-ingestive effects of energy and protein and associate those effects with a food’s flavour to modify food choices (J.J. Villalba and F.D. Provenza, personal communication). Lambs acquire a preference for a poorly nutritious flavoured food eaten during intraruminal infusions of energy (starch) or protein (casein), and shortly after an intraruminal infusion of energy or protein (preload), lambs decrease their preferences for the flavour previously conditioned with starch or casein, respectively. Preloads of casein decreased preferences for flavours previously paired with casein, and increased preferences for flavours paired with starch. Preloads of energy had the opposite effect. These results show that lambs discriminate between the post-ingestive effects of energy (starch) and protein (casein) and associate the effects with specific external cues (i.e. added flavours) to regulate macronutrient ingestion. There is thus accumulating evidence that ruminants prefer to avoid a food, with an adequate nutrient balance overall, that delivers different nutrients at different rates and results in temporary imbalances.

Temporal relationships between ingestion and toxicosis The above raises the question about the time-scale of conditioning. Conditioned aversions to foods are unusual in that it is not necessary for the conditioning and the unconditioned stimuli to be very close in time. In contrast to classical conditioning, ingestion of a flavoured food can become associated with abdominal or metabolic discomfort several hours later (Provenza et al., 1993). In acquiring strong preferences for flavoured straw paired with starch, the delay between straw ingestion and starch infusion affects preferences by lambs (Villalba et al., 1999). Lambs that received starch 1 h after beginning to eat straw had lower preferences than lambs that received starch immediately after beginning to eat straw. The shorter the delay between food ingestion and post-ingestive feedback, the stronger the preference, which is why non-structural carbohydrates that are rapidly digested in the rumen condition strong food preference (Dove, 1996). Sheep actively seek plant parts high in soluble carbohydrates (Simpson and Dove, 1994), which leads to improved efficiency of microbial protein synthesis and increased organic matter intake (Dove and Milne, 1994). The stronger the aversion the longer it persists (Kyriazakis et al., 1998). It is necessary, on the one hand, for conditioned taste aversions to persist (other-

Integration of Learning and Metabolic Signals

7

wise they have no function), but also to be flexible (otherwise an animal might be saddled with an unnecessary aversion for the rest of its life).

Maintenance of stable rumen function The ruminant must maintain a stable rumen environment if the incumbent microbes are not to be threatened. Of particular importance is the regular intake of fibrous material and it is relevant to consider to what extent ruminants exhibit an ‘appetite’ for fibre. Given free choice between forage and concentrates, cattle and goats take about 20% of their DM intake as forage. Campion and Leek (1997) found that sheep offered a diet free of long fibre ate significant amounts of fibre when available separately, and that a polyethylene fibre pompom introduced via the rumen cannula and placed so as to brush the reflexogenic areas of the reticulorumen, reduced voluntary polyethylene fibre intake. Indeed, sheep reduced their intake of a high-energy food, but maintained their intake of a low-energy food when infused with acid or alkali to increase rumen osmolality. It was concluded that the sheep were responding in order to maintain the stability of conditions in the rumen (Engku Azahan and Forbes, 1992; Cooper et al., 1995). Further, Cooper et al. (1996) found that inclusion of sodium bicarbonate in the diet of sheep given a choice between foods with low and high energy density increased the proportion of high-energy feed selected – the bicarbonate ameliorated the low pH induced by the rapid fermentation of the high-energy feed. Sheep self-regulate intake of sodium bicarbonate when fed diets varying in grain content (Phy and Provenza, 1998). These observations are all consistent with a desire by ruminants to maintain a certain degree of fibrousness in their diet in order to stabilize conditions in the rumen.

Trade-offs There are many situations in which the animal must balance its intake or choice of foods in order to trade off the intake of a toxin against the need to obtain nutrients (bearing in mind that, in general, the only difference between a toxin and a nutrient is the concentration in the diet). For example, lambs preferred barley to lucerne pellets in the absence of added toxin (LiCl), but this preference was reversed the higher the content of LiCl in the barley (Wang and Provenza, 1997). Intake of foods containing a constant concentration of LiCl increased as the proportion of barley in the food increased, i.e. animals were better able to tolerate a higher concentration of toxin the higher the yield of energy per kg. Likewise, sheep fed diets high in sagebrush, which contains various terpenes, eat more sagebrush as the macronutrient content of their diet increases (R.E. Banner, J. Rogosic, E.A. Burritt and F.D. Provenza, personal communication). When this finding is translated to the natural situation, in which both toxin and nutrient contents of plants vary with season and location, it emphasizes how nutrients interact to influence food intake. It also illustrates the ability of the body to integrate signals from nutrients and those from toxins. Many forages contain toxic phenolics but these are sometimes the available plant species with the highest yield of digestible nutrients and grazing animals must trade

8

J.M. Forbes and F.D. Provenza

nutrients off against toxins. When mule deer were offered pairs of foods with different phenolic:digestible energy ratios they ate high-energy foods when these were low in phenolics, but low-energy foods when the high-energy ones were high in phenolics (McArthur et al., 1993). Similarly, goats chose to eat more low-digestibility blackbrush twigs when the high-digestibility twigs were high in toxin (either naturally-occurring condensed tannins or added LiCl) (Provenza et al., 1994a). Also, lambs preferred foods higher in readily-available carbohydrate but when a high-energy food had tannin added to it then preference switched to the lower-energy food (C.H. Titus, F.D. Provenza, E.A. Burritt, A. Perevolotsky and N. Silanikove, personal communication). When polyethylene glycol (PEG) was given each morning lambs also shifted their preferences to lower-energy foods as tannin levels increased, but to a significantly lesser extent than their counterparts which did not receive the PEG supplement. Those supplemented with PEG ate more than unsupplemented animals of the tannin-containing foods, particularly as the abundance of nutritious alternatives diminished. Hutchings et al. (1999) have shown how lambs trade off high nutritive value of herbage against the risk of being parasitized with Ostertagia; animals which had been feed-restricted took greater risks of parasitism in order to obtain more nutritious grass than unrestricted lambs. It can be concluded, therefore, that diet selection is directed towards stabilizing conditions in the rumen and the rest of the body, i.e. avoiding disease, metabolic imbalance and upset. It is also presumably directed against over-stimulation of the rumen by physical means, although an attempt to condition a flavour aversion by inflation of a balloon in the rumen of cows did not succeed (Klaiss and Forbes, 1999).

Afferent pathways for conditioning stimuli This section presents evidence that signals from metabolic receptors converge with those from other classes of receptor, such as those monitoring visceral distension, osmolality and adiposity, to provide the CNS with a composite picture of comfort/ discomfort and thus to play a vital part in influencing preferences for foods. This will be done mainly by reference to reviews rather than a detailed description.

Visceral organs From the mouth to the anus, food and digesta provide a continuously changing set of stimuli to stretch and tactile receptors. The complexity of the information provided to the CNS is formidable even though there is considerable convergence in the afferent branches of the autonomic nervous system which relays the impulses (Forbes, 1996). The seminal work of Leek and colleagues, summarized by Forbes and Barrio (1992), showed how physical and chemical stimulation of receptors in the digestive tract activates vagal afferent pathways to the gastric centre of the medulla oblongata. In addition, the liver is sensitive to propionate, once again relaying its information to the CNS via autonomic afferents (Anil and Forbes, 1987) and providing a comprehensive assessment of the energy supply via the liver. If a simple experimental procedure such as inflating a balloon in the rumen, or

Integration of Learning and Metabolic Signals

9

infusing the salt of a volatile fatty acid at a constant rate, has complex effects then how much more complex are the effects of the passage of a meal through the digestive tract, with ever-changing physical and chemical properties? It seems unlikely that the CNS could interpret individually the nature and degree of stimulation of each receptor as there is both polymodal and polytopic integration of visceral signals with convergence in hind, mid and fore brains (Forbes, 1996). Thus, severe stimulation of receptors in a specific part of the digestive tract is felt as a dull, mid-line sensation by humans so presumably milder stimulation by physiological stimuli are also not ascribed by the CNS to specific sites in the abdomen. It is not necessary to be consciously aware of abdominal discomfort in order for the CNS to be able to respond to events in gut and liver, i.e. to participate appropriately in the control of gut motility and secretion, and to remember the sensory properties of the food eaten shortly before the discomfort occurred (Provenza et al., 1994b; Forbes, 1998). Abdominal pain is likely to be an extension of normal sensitivity to stimuli such as distension of a viscus or irritation of a mucosal surface by chemicals so that metabolic imbalance is relayed as a mild form of toxicity.

Body tissues Some parts of the CNS, such as the nucleus of the solitary tract (NTS) and hypothalamus, are sensitive to a shortage of available energy (Grill, 1986). Starving cells of glucose by treatment with 2-deoxy-D-glucose, which blocks glucose uptake by cells, has shown this. An excess of glucose in the blood does not influence the NTS activity, only a shortage. For the rest of the body to fail to protect the CNS from starvation must be seen as an unusual and serious situation and one that would be likely to be a potent conditioning stimulus to eat more food in future. At a broader level, adipose tissue has recently been shown to produce a hormone, leptin or ob-protein, increasingly as adipocyte size increases. Leptin is taken up in the CNS where it inhibits feeding via the neuropeptide Y system (Houseknecht et al., 1998). It is reasonable to assume, for the time being, that any negative feedback due to leptin is integrated with stimuli emanating from the viscera so that the normal meal-tomeal regulation is modulated to a slight extent, but one which, being persistent, results in significant long-term reduction in intake in obese animals. We see no need to consider short-term and long-term controls of intake as being different in nature, just in the rate of change of the signals involved. No muscle hormone equivalent to leptin has so far been found and it seems likely that protein deposition exerts its effect on feeding by taking amino acids out of the circulation and thus influencing liver metabolism.

‘Metabolic discomfort’ as a concept to describe the sensations relayed to the CNS from receptors sensitive to supply and balance of nutrients We propose, therefore, that there is a complex set of signals coming from many parts of the body, the strength of which is dependent on previous nutrition (i.e. feeding behaviour), both short- and long-term. Short-term excesses of nutrient intake result in

10

J.M. Forbes and F.D. Provenza

over-stimulation of gut and liver receptors, and long-term excesses result in over-production of leptin, which generate a state of ‘metabolic discomfort’, otherwise known as ‘satiety’. Conversely, under-stimulation of liver and hind-brain receptors due to undereating (relative to demands) generates a different type of metabolic discomfort, especially when coupled with low levels of leptin, usually known as ‘hunger’. We view the range from extreme hunger to excessive satiety as a continuum. The animal directs its behaviour to achieving the most ‘comfortable’ situation, in which comfort is not only induced by the appropriate supply of nutrients, but also by social pressures to eat or not to eat, and by learned associations (see above). The anatomical and physiological mechanisms underlying affective and cognitive systems have been fairly well established (Provenza et al., 1998). Taste afferents converge with visceral afferents in the solitary nucleus of the brain stem and proceed to the limbic system, where the hypothalamus and related structures maintain homeostasis in the internal environment through the endocrine system, the autonomic nervous system, and the neural system concerned with motivation and drive (i.e. incentive modification). Higher cortical centres interact with the hypothalamus through the limbic system, and regulate the internal environment primarily by indirect action on the external environment (i.e. behaviour modification). These alternative means of regulating the internal environment generally function in parallel. For example, the taste of food is adjusted according to that food’s effect on the internal environment; on this basis, animals use thalamic and cortical mechanisms to select foods that are nutritious and avoid those that are toxic. Given a choice of foods, and the ability to learn to eat that ratio between the foods which balances nutrient supply with demand, we interpret the evidence presented above to suggest that animals adjust their food choice to minimize metabolic discomfort. Thus, in meeting its ‘requirements’ it is not fulfilling some long-term goal, but rather reacting to hard- and soft-wired programmes, i.e. signals from body organs and tissues and the associations previously established between these signals and the sensory properties of the food (Provenza et al., 1998).

If nutrient requirements change then level of comfort changes and choice is adjusted to regain a state of maximum comfort If it is true that the animal eats to minimize discomfort then any change in the nutrient uptake by tissues should lead to a change in the selection made between a choice of foods. Protein requirements are higher for pregnant than for non-pregnant ewes and when Cooper et al. (1994) offered non-pregnant and late-pregnant ewes the choice between high (HP) and low-protein (LP) foods, both of high energy content, they found that the pregnant animals selected a significantly greater proportion of HP than those not pregnant. Lactating cows offered free access to grass silage and a restricted amount of concentrates, offered as a choice one with greater (HP) and another with less (LP) than the estimated requirement for MP, ate a greater proportion of HP the higher their output of milk protein (Lawson et al., 1999), i.e. they appeared to be directing their selection between foods to supply their ‘requirements’ for protein.

Integration of Learning and Metabolic Signals

11

Growing lambs chose a diet well-matched to their requirements for growth (Hou et al., 1991a) and worked to maintain this balanced diet when one of the foods required up to 30 responses to obtain a reinforcement (Hou et al., 1991b). Thus, diet selection is not only driven by the composition of the foods on offer, but also by the requirements of the animal, which change in a systematic manner with reproductive cycles.

What is true for choice between foods is, we propose, also true for a single food An animal ‘experiments’ with different levels of intake of a single food until its comfort levels are optimized In much of the previous discussion we have concentrated on the control of dietary choice, but we now propose that a logical extension of the proposal that dietary choice is directed to minimizing metabolic discomfort, is that the intake of a single food is also directed to achieving the same end. If our hypothesis is true then we should be able to find evidence that animals given a single food ‘experiment’ by increasing and decreasing their intake in a cyclical manner in order to continually ascertain whether their comfort is maximized. Cyclical intake of foods containing toxins Such cyclicity might be most easily seen with a food containing an overt toxin such as tall larkspur. Pfister et al. (1997) showed that cattle given pelleted food containing 27% larkspur ad libitum showed distinct cyclic patterns of intake, where increased larkspur consumption on 1 or 2 days was followed by reduced consumption on the following day. The amount of larkspur (mean 2.0 kg day⫺1; 17.8 mg toxic alkaloid kg⫺1 body weight) consumed was just below a level that would produce overt signs of toxicity. Further experiments with LiCl and larkspur given in a maize supplement confirmed this cyclical pattern of intake. Evidence of cyclicity for ‘normal’ foods Feeding behaviour is, by its very nature, cyclical and examination of daily feed intakes of individual animals shows considerable day-to-day variation, sometimes by as much as twofold, too large to be due simply to any changes in the environment or the physical and chemical properties of the forage. An animal that eats exactly the same amount of food each day has no opportunity to find out whether its overall comfort would be better served by eating a little more, or a little less, than on the previous days. External factors affect food intake, including the composition of the food, climatic variables and changes in management procedures. These include variations in the time of day at which the day’s refusals are measured and the time spent without access to food at milking time or when fresh food is being delivered. Such variations in daily intake, if they can be monitored and associated with changes in feedback from visceral and other receptors, could allow the animal to assess whether it should eat more or less than before. However, the effects of such variables on daily intake should be similar for all animals. The extent to which there is endogenous

12

J.M. Forbes and F.D. Provenza

variation can be investigated by examining the residual variation from regression of daily intake of one animal’s intake against the mean intake of the rest of the group. A difficulty in describing day-to-day variation in intakes is that few papers, if any, publish daily intakes of individuals. We therefore obtained unpublished daily intakes of a maize silage/concentrate mixed feed by 12 lactating cows over a 31-day period (EMBRAPA, Juiz de Fora, Brazil, 1999, unpublished results). There was an approximately 5-day cycle in the mean intake of the group but the intakes of individuals show some considerable additional variation. The mean standard deviation of the residuals for the linear regressions of the intakes of each individual against the group intake for this data set was 3.63 with a range 2.00 to 5.13 kg day⫺1, i.e. there is considerable individual variation in addition to the variation in daily intake shown by the whole group. Anyone who works with animals knows that voluntary intake by individuals varies day by day. An hypothesis to account for cyclicity When given a chance, animals eat a variety of foods and they forage in many different locations. We contend that animals satiate (or get bored) from eating the same foods in the same places (Provenza, 1996). The more adequate the food(s) or location(s) nutritionally, the less animals satiate, but they satiate none the less. For example, cattle fed either plain or ammoniated straw in one flavour or another (either maple or coconut) for 1 day strongly preferred the alternate flavour when offered a choice for 2 h day⫺1 for the next 5 days. The change in preference was stronger for plain than for ammoniated straw (S.B. Atwood, F.D. Provenza, R.D. Wiedmeier and R.E. Banner, personal communication). For similar reasons, sheep prefer foods in a variety of flavours and nutrient contents (Early and Provenza, 1998; J.J. Villalba and F.D. Provenza, personal communication), and they forage in locations with a variety of foods (Scott and Provenza, 1998). The same is true for toxins: the higher the toxin content of a food, the more readily animals satiate on the food (E.A. Burritt and F.D. Provenza, personal communication). Thus, satiety causes animals to eat different foods and to forage in different locations. Access to a single food induces boredom for that food and a reduction in intake, followed by hunger and increased intake; hence the cyclical pattern of intake observed. Variety is the spice of life because bodies satiate on familiarity and thrive on variety. In a changing world, it is the body’s way to encourage old (sheep)dogs to learn new tricks.

A change in the composition of the food, or in the requirements for one or more nutrients, results in a gradual re-learning of the amount to be eaten to re-establish optimum comfort Changes in food composition lead to changes in food intake. If the change is in the nutritional value then the animals may well find that a higher or lower daily dry matter (DM) intake gives it more comfort than simply continuing to eat the same daily weight as before the change. On the other hand, if the sensory properties of the food change without significant change in nutrient yield any change in intake level is usually shortlived. Conversely, a change in flavour causing a temporary drop in acceptability can be masked by the inclusion of a familiar flavour (Frederick et al., 1988).

Integration of Learning and Metabolic Signals

13

Change of diet A particular situation in which it is necessary for animals to learn about food is when the food to which they have been accustomed is suddenly changed to another with different sensory and nutritional properties. In most practical and experimental situations sudden changes are avoided by a slow change of diet; in changeover experiments it is usual to wait for animals to stabilize on a new diet before recommencing recording of feeding behaviour. Mean daily intakes of 32 individual sheep accustomed to oaten chaff and then suddenly changed to barley straw are shown in Fig. 1.1, together with intakes of two individuals, selected at random (J. Hills, 1998, unpublished results). On the first day after the change the sheep ate very little which suggests that they were not familiar with the new food and were showing neophobia – a wariness of new food in which only small amounts are taken in order to assess whether there are unpleasant consequences to eating it. A combination of increasing hunger and the realization that the new food does not cause illness encouraged increased acceptance of the food and a steadily increasing intake. On the 4th and 5th days there is a decline in intake, perhaps due to rumen disturbance if the microflora has not adapted to the new food, followed by a gradual climb to a plateau, stabilizing some 10 days after introduction of the barley straw. The two individuals shown on Fig. 1.1 conform quite well to the mean for all 32 animals, but with greater day-to-day fluctuations than the mean. When the more stable data from the last 15 days were analysed, as for the cow data above, there were again large day-to-day fluctuations in intake by individuals, in comparison with fluctuations in the mean for all animals. The mean standard deviation of the residuals was 82, with a range from 31 to 162 g day⫺1. Only in 15 out of the 32 individual sheep was daily intake in the last 15 days significantly (P < 0.05) correlated with the daily mean intake of all 32 sheep. For the two individuals in Fig. 1.1, the one represented by the dashed line was significantly correlated, but not the other. This

Fig. 1.1. Mean daily intakes for 32 sheep (solid line) and two individual sheep (lines with symbols) selected at random; on 6 September the oaten chaff was replaced with barley straw (J. Hills, unpublished results).

14

J.M. Forbes and F.D. Provenza

again is clear evidence for large variations in daily intake that are unrelated to external factors such as food quality or climatic variables. Even if this day-to-day variation is not purposeful, it still serves the animals’ intake control system by providing information about whether a small increase or decrease in intake improves metabolic comfort. Changes in the animal After weaning, the intake of DM usually increases in proportion to energy requirements for growth. As maturity approaches and the amount of fat in the body increases, food intake reaches a plateau, although in many ruminants there is an annual cycle of intake and body fat content (Forbes, 1995). These changes occur slowly and there is plenty of time for animals to learn whether small increases or decreases in intake are necessary to conserve minimal discomfort. On the other hand, as a cow progresses from late pregnancy into early lactation there are much more rapid physical and metabolic changes including a reduction in the competition for abdominal space, changes in amounts and ratios of nutrients required, oestrogens increasing in late pregnancy and then falling at parturition, and the distraction of parturition itself. Thus, there is a whole new balance of factors required in order to achieve optimal metabolic comfort which takes the animal a long time to get right, hence the slow increase in early lactation.

Minimal total discomfort – a modelling approach There are numerous models of voluntary food intake by ruminants (see Chapter 2), but integration of metabolic, physical, social and other factors, where attempted, has been dealt with somewhat superficially. Although we have advocated the concept of additivity of different feedback factors, we are not aware of any models that have properly taken this into account (but see Fisher, 1996), although some have acknowledged the need to do so (Poppi et al., 1994). Our thesis is that signals from the various families of visceral receptors (including signals generated by numerous metabolites interacting at the level of the liver and other organs that are sensitive to metabolic fluxes), together with signals from adipose tissue, social stimuli and environmental factors, are integrated by the CNS in an approximately additive manner to generate a total signal of ‘discomfort’. The animal then adjusts its intake and/or choice, continuing in the direction that results in a reduction in discomfort and learning as it goes. In a complex version of such a model there would be representation of numerous nutrients and metabolites as well as a detailed description of physical effects in the digestive tract and external factors such as social and environmental effects. The relationship between the strength of each of these signals and the frequency of impulses in afferent nerves would need to be known and such information is not available in most cases. For purposes of the simple model presented here we have taken ME and crude protein (CP) to represent important classes of nutrients. We assumed that the effect of a deviation in nutrient supply from the food from the animal’s ‘requirement’ has an effect on comfort which is proportional to the square of the deviation (this has the advantage of getting rid of negatives and also emphasizes large deviations from optimal). Therefore a deficiency of a nutrient has the same unbalancing effect as an

Integration of Learning and Metabolic Signals

15

excess. A third factor affecting intake in this model is neutral detergent fibre (NDF), as an index of the bulk of the food and thus rumen fill. Fibre is included because an excess inhibits food intake while the animal’s digestion suffers if there is a deficiency, in a manner parallel to the way that energy and protein supply are proposed to generate discomfort when either greater or less than the rate at which they are being utilized by the animal. Fig. 1.2 gives an example in which discomfort is plotted against intake of a typical forage. This illustrates how the sum of the discomforts due to several properties of the food, in relation to the ‘requirements’ of the animal, generates a total discomfort signal which the animal attempts to minimize, according to the hypotheses presented here. Iteration proceeds until the intake that produces the minimum metabolic discomfort is reached. Forbes (1999) describes the model in further detail. When the ME, CP and NDF contents of the foods used by Kyriazakis and Oldham (1993) are used as parameters in the model, together with the requirements of their lambs, good agreement of daily intake was obtained for the high- and mediumprotein foods. Predicted intakes of the low-protein foods were much greater than observed, however, suggesting that greater weighting be put on protein deficiency as a contributor to metabolic discomfort and no doubt the weighting of the signals from deviations in ME, CP and NDF will need refinement. When two foods with different composition are available the model ‘experiments’ by changing the rates of intake of each until minimum discomfort is reached, thereby predicting both daily food intake and the proportion of each of the two foods eaten. The choices made by sheep when offered two foods with different protein contents in the same experiment as that discussed in the previous paragraph (Kyriazakis and Oldham, 1993) were modelled. Where possible the lambs chose foods in a ratio that

Fig. 1.2. Discomfort due to: ⽧, metabolizable energy (ME); 䡲, crude protein (CP); neutral detergent fibre (NDF); ⴒ, total; plotted against food intake. Discomfort is the square of the weighted proportional deviation of the supply of each ‘nutrient’ from the animal’s ‘requirement’. In this example the requirements were those of a growing lamb for 20 Mj ME, 250 g CP and 350 g NDF day⫺1 and the forage food provided 10 Mj ME, 120 g CP and 600 g NDF kg⫺1. Given the hypothesis that the animal eats that amount of food at which total discomfort is minimized, the predicted intake in this example is 1.2 kg day⫺1.

䊱,

16

J.M. Forbes and F.D. Provenza

gave approximately an optimum protein content in the diet while the model provides a ‘perfect’ dietary protein content and minimal ‘metabolic discomfort’ with choice feeding. Choice-fed lambs were observed to eat more in total than similar sheep offered single foods while the model predicts daily intakes of single- and choice-fed animals to be similar. Thus, the model behaves in a similar way to animals in some respects but not in others. A number of improvements are immediately apparent. The model could be refined mathematically to avoid iteration in arriving at a solution; however, this iteration does, according to our hypothesis, mimic what is going on in the animal. Many more qualities of the food (nutrients) should be included once there is a better understanding of how they are involved in pathways that contribute to metabolic comfort/discomfort. Feedbacks from body reserves (insulin, leptin) should be included as signals added to those emanating from receptors in the digestive tract and liver to incorporate the long-term aspects of intake control. We believe, however, that this novel approach to simulating and predicting intake and choice by ruminants and other animals allows a flexibility not previously possible in incorporating whatever level of detail in metabolism is required by the user – the form of the model presented here is the simplest possible.

Conclusions and future research A very important task of the ruminant nutritionist is to minimize the cost of a diet (by including cheap, fibrous materials) while ensuring adequate intake of digestible nutrients. We need to know more about how ruminants trade off the various factors tending to stimulate and inhibit feeding, especially the balance between nutrients and fibre. We do not envisage any demarcation between ‘short-term’ and ‘long-term’ controls of intake or diet selection. It is important, therefore, to establish experimentally whether there is additivity between effects of rapidly-changing visceral signals, such as distension and volatile fatty acids, and slow-changing indicators of adiposity such as leptin. Experimental approaches need to be devised to discover whether animals can remember how much they have eaten in a day. Can they integrate feedback signals over a whole day? 24 h seems to be a sensible time over which to integrate knowledge about food and its effects as there are diurnal rhythms of feeding due to the dark–light cycle, and the regular times of feeding and milking in housed animals. Different individuals meet their similar requirements by very different patterns of feeding behaviour within the day (Tolkamp and Kyriazakis, 1997). There is also considerable variation between individuals in their choice between foods and, as outlined above, in the intake in an individual on consecutive days. Such variations tend to have been overlooked with the advent of statistics during the 20th century which has advanced our ability to design and analyse experiments, and thus to better understand biological processes, but has obscured the vital importance of variation among individuals in biological processes (Provenza et al., 1999). In studies of behaviour and nutrition we typically determine nutritional needs and formulate rations for the average member of the herd and not for specific individuals. Nevertheless, variation among individuals could affect productivity of the group if the ration diverges too much from

Integration of Learning and Metabolic Signals

17

what individuals at the extremes – which can be as much as half of the group (Provenza et al., 1996) – prefer and can tolerate. Diets that enable animals to select among foods may better enable each individual to best meet its nutritional needs, and under some circumstances, may lower daily feed costs. Above all, we emphasize the need to acknowledge that metabolic factors, physical factors and learning all have important roles to play in the complexities of the control of food intake in ruminant animals. We need to recognize the true multifactorial nature of the control of voluntary intake and diet selection if we are to advance understanding and predictive ability.

Acknowledgements The authors are grateful to EMBRAPA, Juiz de Fora, Brazil, and to James Hills, University of New England, Australia, for access to their unpublished data of daily food intakes of individual animals; also to Ilias Kyriazakis for critically reviewing the manuscript.

References Anil, M.H. and Forbes, J.M. (1987) Neural control and sensory functions of the liver. Proceedings of the Nutrition Society 46, 125–133. Arsenos, G. and Kyriazakis, I. (1999) The continuum between preferences and aversions for flavoured foods in sheep conditioned by administration of casein doses. Animal Science 68, 605–616. Burritt, E.A. and Provenza, F.D. (1992) Lambs form preferences for nonnutritive flavors paired with glucose. Journal of Animal Science 70, 1133–1136. Campion, D.P. and Leek, B.F. (1997) Investigation of a ‘fibre appetite’ in sheep fed a ‘long fibrefree’ diet. Applied Animal Behaviour Science 52, 79–86. Cooper, S.D.B., Kyriazakis, I. and Oldham, J.D. (1994) The effect of late pregnancy on the diet selections made by ewes. Livestock Production Science 40, 263–275. Cooper, S.D.B., Kyriazakis, I. and Nolan, J.V. (1995) Diet selection in sheep – the role of the rumen environment in the selection of a diet from 2 feeds that differ in their energy density. British Journal of Nutrition 74, 39–54. Cooper, S.D.B., Kyriazakis, I. and Oldham, J.D. (1996) The effects of physical form of feed, carbohydrate source, and inclusion of sodium-bicarbonate on the diet selections of sheep. Journal of Animal Science 74, 1240–1251. Denton, D.A. (1982) The Hunger for Salt. Springer Verlag, Berlin. Dove, H. (1996) Constraints to the modeling of diet selection and intake in the grazing ruminant. Australian Journal of Agricultural Research 47, 257–275. Dove, H. and Milne, J.A. (1994) Digesta flow and rumen microbial protein production in ewes grazing perennial ryegrass. Australian Journal of Agricultural Research 45, 1229–1245. Dutoit, J.T., Provenza, F.D. and Nastis, A. (1991) Conditioned taste-aversions – how sick must a ruminant get before it learns about toxicity in foods. Applied Animal Behaviour Science 30, 35–46. Early, D. and Provenza, F.D. (1998) Food flavor and nutritional characteristics alter dynamics of food preference in lambs. Journal of Animal Science 76, 728–734. Engku Azahan, E.A. and Forbes, J.M. (1992) The effects of intraruminal infusions of sodium salts on the selection of hay and concentrate foods by sheep. Appetite 18, 143–154.

18

J.M. Forbes and F.D. Provenza Fisher, D.S. (1996) Mathematical integration of distension and chemostatic feedbacks for regulation of ruminant intake in diets with varied dietary protein. Journal of Animal Science 74, 3076–3081. Forbes, J.M. (1995) Voluntary Food Intake and Diet Selection in Farm Animals. CAB International, Wallingford, 520 pp. Forbes, J.M. (1996) Integration of regulatory signals controlling forage intake in ruminants. Journal of Animal Science 74, 3029–3053. Forbes, J.M. (1998) Dietary awareness. Applied Animal Behaviour Science 57, 287–297. Forbes, J.M. (1999) Minimal total discomfort as a concept for the control of food intake and selection. Appetite 33, 371. Forbes, J.M. and Barrio, J.P. (1992) Abdominal chemo- and mechanosensitivity in ruminants and its role in the control of food intake. Experimental Physiology 77, 27–50. Frederick, G., Forbes, J.M. and Johnson, C.L. (1988) Masking the taste of rapeseed meal in dairy compound food. Animal Production 46, 518. Grill, H.J. (1986) Caudal brainstem contributions to the integrated neural control of energy homeostasis. In: Ritter, R.C., Ritter, S. and Barnes C.D. (eds) Feeding Behavior – Neural and Humoral Controls. Academic Press, New York, pp. 103–129. Hills, J., Nolan, J.V., Hinch, G.N. and Lynch, J.J. (1998) Selection of diets differing in sulphur content by sheep in sulphur adequate or depleted states. Proceedings of the British Society of Animal Science 1998, 28. Hou, X.Z., Emmans, G.C., Anderson, D., Illius, A.W. and Oldham, J.D. (1991a) The effect of different pairs of feeds offered as a choice on food selection by sheep. Proceedings of the Nutrition Society 50, 94A. Hou, X.Z., Lawrence, A.B., Illius, A.W., Anderson, D. and Oldham, J D. (1991b) Operant studies on feed selection in sheep. Proceedings of the Nutrition Society 50, 95A. Houseknecht, K.L., Baile, C.A., Matteri, R.L and Spurlock, M.E. (1998) The biology of leptin: A review. Journal of Animal Science 76, 1405–1420. Hutchings, M.R., Kyriazakis, I., Gordon, I.J. and Jackson, F. (1999) Trade-offs between nutrient intake and faecal avoidance in herbivore foraging decisions: the effect of animal parasitic status, level of feeding motivation and sward nitrogen content. Journal of Animal Ecology 68, 310–323. Klaiss, M. and Forbes, J.M. (1999) Effect of conditioning feed flavour to rumen distension in cows. Proceedings of the British Society of Animal Science, p. 85. Kyriazakis, I. and Oldham, J.D. (1993) Diet selection in sheep: the ability of growing lambs to select a diet that meets their crude protein (nitrogen ⫻ 6.25) requirements. British Journal of Nutrition 69, 617–629. Kyriazakis, I. and Oldham, J.D. (1997) Food intake and diet selection in sheep: The effect of manipulating the rates of digestion of carbohydrates and protein of the foods offered as a choice. British Journal of Nutrition 77, 243–254. Kyriazakis, I., Anderson, D.H. and Duncan, A.J. (1998) Conditioned flavour aversions in sheep: the relationship between the dose rate of a secondary plant compound and the acquisition and persistence of aversion. British Journal of Nutrition 79, 55–62. Lawson, R.E., Redfern, E.J. and Forbes, J.M. (2000) Choices by lactating cows between concentrates high or low in digestible undegraded protein. Animal Science (in press). McArthur, C., Robbins, C.T., Hagerman, A E. and Hanley, T.A. (1993) Diet selection by a ruminant generalist browser in relation to plant chemistry. Canadian Journal of Zoology 71, 2236–2243. Pfister, J.A., Provenza, F.D., Manners, G.D., Gardner, D.R. and Ralphs, M.H. (1997) Tall larkspur ingestion – Can cattle regulate intake below toxic levels? Journal of Chemical Ecology 23, 759–777. Phy, T.S. and Provenza, F.D. (1998) Eating barley too frequently or in excess decreases lambs’

Integration of Learning and Metabolic Signals

19

preference for barley but sodium bicarbonate and lasalocid attenuate the response. Journal of Animal Science 76, 1578–1583. Poppi, D.P., Gill, M. and France, J. (1994) Integration of theories of intake regulation in growing ruminants. Journal of Theoretical Biology 167, 129–145. Provenza, F.D. (1996) Acquired aversions as the basis for varied diets of ruminants foraging on rangelands. Journal of Animal Science 74, 2010–2020. Provenza, F.D., Nolan, J.V. and Lynch, J.J. (1993) Temporal contiguity between food ingestion and toxicosis affects the acquisition of food aversions in sheep. Applied Animal Behaviour Science 38, 269–281. Provenza, F.D., Lynch, J.J., Burritt, E.A. and Scott, C.B. (1994a) How goats learn to distinguish between novel foods that differ in postingestive consequences. Journal of Chemical Ecology 20, 609–624. Provenza, F.D., Lynch, J.J. and Nolan, J.V. (1994b) Food aversion conditioned in anesthetized sheep. Physiology and Behavior 55, 429–432. Provenza, F.D., Scott, C.B., Phy, T.S. and Lynch, J.J. (1996) Preference of sheep for foods varying in flavors and nutrients. Journal of Animal Science 74, 2355–2361. Provenza, F.D., Villalba, J.J., Cheney, C.D. and Werner, S.J. (1998) Self-organization of foraging behavior: From simplicity to complexity without goals. Nutrition Research Reviews 11, 199–222. Provenza, F.D., Villalba, J.J. and Augner, M. (1999) The physics of foraging. In: BuchananSmith, J.G., Bailey, L.D. and McCoughey, P. (eds) Proceedings of the XVIII International Grassland Congress. Extension Service, Saskatchewan Agriculture and Food, Saskatoon, Saskatchewan, pp. 99–107. Scott, L.L. and Provenza, F.D. (1998) Variety of foods and flavors affects selection of foraging locations by sheep. Applied Animal Behaviour Science 61, 113–122. Simpson, R.J. and Dove, H. (1994) Plant non-structural carbohydrates, diet selection and intake. Proceedings of the Australian Society of Animal Production 20, 59–61. Tolkamp, B.J. and Kyriazakis, I. (1997) Measuring diet selection in dairy cows: effect of training on choice of dietary protein level. Animal Science 64, 197–207. Villalba, J.J. and Provenza, F.D. (1996) Preference for flavored wheat-straw by lambs conditioned with intraruminal administrations of sodium propionate. Journal of Animal Science 74, 2362–2368. Villalba, J.J. and Provenza, F.D. (1997a) Preference for flavored wheat straw by lambs conditioned with intraruminal infusions of acetate and propionate. Journal of Animal Science 75, 2905–2914. Villalba, J.J. and Provenza, F.D. (1997b) Preference for wheat straw by lambs conditioned with intraruminal infusions of starch. British Journal of Nutrition 77, 287–297. Villalba, J.J., Provenza, F.D. and Rogosic, J. (1999) Preference for flavored wheat straw by lambs conditioned with intraruminal infusions of starch administered at different times after straw ingestion. Journal of Animal Science 77, 3185–3190. Wang, J. and Provenza, F.D. (1997) Dynamics of preference by sheep offered foods varying in flavors, nutrients, and a toxin. Journal of Chemical Ecology 23, 275–288.

2

Mathematical Models of Food Intake and Metabolism in Ruminants A.W. ILLIUS1, N.S. JESSOP1 AND M. GILL2 1Division 2NR

of Biological Sciences, University of Edinburgh, Edinburgh, UK; International, Central Avenue, Chatham Maritime, Kent, UK

Introduction The daily rate of food intake is the single most important factor affecting animal performance and productive efficiency. Knowledge of food intake is necessary for diet formulation, for the prediction of animal performance, for the design and control of production systems and for the assessment of animal–resource interactions in grazing ecosystems. Modelling the control of food intake and nutrient supply is a way of furthering our understanding of mechanisms, of testing the consequences of our assumptions, and of developing a mechanistic framework capable of accurate prediction. There are, broadly, two types of models: ‘digesta kinetics’ and ‘metabolic’ models. The first type is concerned with the prediction of intake and digestion, on the assumption that rate of intake is limited by the rate of decrease in volume of rumen digesta by digestion and passage. The second type is concerned with the production, absorption and utilization of nutrients via microbial and animal metabolism. Whilst the former type often relies on empirical estimates of parameters, and is mostly concerned with prediction of intake and digestion, the latter type of model is rarely used to predict intake, and indeed generally requires intake as a model input. Instead, it is most often used in the pursuit of knowledge. We will exclude from this review statistical models used for data analysis. The distinction between these types of model goes back to the early 1970s, when two models were published that had a profound influence on subsequent modelling: Baldwin et al. (1970) and Waldo et al. (1972). These models are very different in complexity and represent divergent schools of thought. Baldwin et al. (1970) modelled the rumen with chemically defined substrates and emphasized the stoichiometry of fermentation and the prediction of fermentation end-products. Digestion was a second-order process affected by microbial mass. This model influenced many subsequent metabolic models which describe the process of digestion as the appearance of fermentation endproducts, or are purely concerned with intermediary metabolism (Gill et al., 1984). © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

21

22

A.W. Illius et al.

The simple model of Waldo et al. (1972) was concerned with cellulose disappearance rather than appearance of fermentation end-products, and partitioned cellulose into a fraction that is potentially digestible by rumen microbes and an indigestible fraction. The substrate was thus defined biologically rather than chemically, and the rates of digestion and passage were assumed to be first order. At present, the major uncertainty in the prediction of intake lies less in the cases where intake is constrained by low digesta clearance rate than where voluntary intake is limited by feedback from metabolic factors. Thus, variation in daily intake can more clearly be related to the kinetics of digestion and passage, assessed on daily timescales, than to shorter-term metabolic events. Voluntary intake is reduced by low diet digestibility and passage rate, and it is therefore assumed that low rates of ruminal digestion and passage lead to physical limitations on daily intake (Laredo and Minson, 1973). More highly digestible feeds can, potentially, be eaten in greater quantities before the presumed physical constraints of gut clearance apply, and voluntary intake is then more likely to be determined by metabolic constraints which are related to the animal’s ability to utilize absorbed nutrients. The role of metabolic constraints in controlling intake of low-quality forages has been demonstrated (Egan, 1977), resulting in reappraisal of whether physical constraints are the major determinant of intake of such forages (Tolkamp and Ketelaars, 1992; Weston, 1996; Pitroff and Kothman, 1999). Although the interaction between physical and metabolic factors is of evident interest and importance for our understanding of the control food intake, there have been few attempts to address it using modelling, perhaps because of the different time-scales and data derivation methods adopted between the two modelling approaches outlined above. This chapter reviews the main themes and issues in modelling the processes governing food intake in ruminants. It addresses the two broad objectives of modelling: making predictions and pursuing knowledge. An important application is to make accurate predictions of intake and digestion for specified combinations of food and animal characteristics. Ruminant research is, in part, conducted in order that livestock industries and society will benefit, and models are an effective means of consolidating and applying diverse information from ruminant research programmes. However, modelling is also an important investigative procedure in its own right. It allows us to investigate mechanisms, test the consequences of our assumptions and hypotheses, show where knowledge or data are lacking, and may be used to define hypotheses and treatments in an experimental programme generating new data.

Review of the task facing modellers Modellers aim to represent the phenomenon of interest in mathematical form. The simplest phenomenological description is a regression model, in which variance in intake is accounted for by independent variables without their necessarily having a causal relationship with intake. These empirical relationships between feed factors and voluntary intake have the merit that they are the best-fit explanations of variation in intake. Depending on how much of the variance they explain, they may be a sound basis for prediction provided that extrapolation beyond the range of the independent variables is avoided. Phenomenological and statistical models should aim to describe

Mathematical Models of Food Intake and Metabolism in Ruminants

23

the system’s behaviour with the smallest number of parameters required to reach the prescribed level of accuracy. Far from attempting a thorough description of the underlying processes, the aim of this type of modelling is to achieve clarity by eliminating unnecessary detail. On the other hand, mechanistic models attempt to represent the underlying causal relationships, and are, in principle, capable of depicting the complexities of the system more faithfully. However, the depth of knowledge across all factors is very uneven, and it should be remembered that increasing complexity is likely to reduce accuracy and tractability. Certainly, any attempt at a complete representation of every causal link is futile and has to be abandoned at some point. At that point, a mechanistic model becomes phenomenological, and relies on empirical relationships. For example, even the most detailed models of rumen function examined by Bannink et al. (1997a) inevitably employ empirical descriptions at some point, for example to describe particle dynamics and feed degradation characteristics. The point is often made (Forbes and France, 1993) that a mechanistic model of some phenomenon occurring at level i is one that may consist of empirical relationships at level i ⫺ 1. Despite what might be regarded as a weakness (Faichney, 1993), the use of empirical descriptions in otherwise mechanistic models should be recognized as a strength: a representation of mechanisms is supported by a robust parameter description. The usual proviso applies about only using empirical estimates within the range of the independent variables used to derive the estimate. Rumen and intake modelling have been pursued by gradually increasing the amount of detailed description of mechanisms, as they become elucidated. But the important question of how much detail is necessary and justified in a model depends on the model’s purpose. Accurate prediction requires robustness, and that requires simplicity and a sound basis for parameter estimation. Exploratory modelling justifies the inclusion of more detail, and the use of parameters whose value cannot readily be estimated. Thus, the parameter values become the experimental variable, and the model’s performance is then related to systematic variation in those values (Bannink et al., 1997a). Inadequate detail in models can lead to erroneous predictions. From consideration only of the degradation rates of feed constituents it was predicted by Sinclair et al. (1993) that asynchrony in the supply of fermentable nitrogen and carbohydrate would cause rumen microbes to be limited by the supply of fermentable nitrogen for much of the day, with negative consequences for the production of fermentation end-products. Empirical results have failed to support this. Sinclair et al. (1993) did not, apparently, consider endogenous contributions (urea recycling, sloughing of cells from rumen epithelium and recycling of microbial matter in the rumen) to fermentable nitrogen. Inclusion of the effect of these endogenous N-sources in a model suggested that ammonia and microbial recycling, and the contribution of hind gut fermentation, would reduce the asynchrony in the balance of substrates available to microbes and of nutrients absorbed into the bloodstream (Illius and Jessop, 1996). The model predicted that microbial production is mostly limited by the rate of supply of fermentable carbohydrate, the converse of that predicted by ignoring supply of endogenous N. In principle, metabolic constraints on food intake result when there is a build-up in concentration of one or more nutrients. This occurs when the rate of metabolism is limited, either by an imbalance of the other metabolites required to synthesize protein or fat, or because the rate is approaching the genetic potential of the animal. This will

24

A.W. Illius et al.

vary with the animal’s stage of growth, reproductive status, physical and climatic environment, and ability to store or otherwise dispose of any surpluses. The balance of nutrients required will depend on the mix of purposes to which they are put, such as for protein or fat deposition, lactation, thermoregulation or locomotion. Some deviation from this ratio can be accepted, provided that body stores can be added to or depleted to balance intake with requirements at maximal production. Surplus amino acids or minerals may also be de-aminated or excreted. Further deviations from the optimal diet, such as would cause a deficit of an essential amino acid relative to the other nutrients absorbed, may cause the animal to compensate by eating more, provided it can dispose of the nutrients in surplus. More extreme dietary imbalances may result in reduced intake. Based on this framework, Illius and Jessop (1996) developed a conceptual model of metabolic constraints on intake of diets varying in energy and protein yield of absorbed nutrients. It is not at all clear how the principles of metabolic control over intake are actually expressed by the animal, in terms of meal size and frequency, and daily intake, although it is clear that cessation of eating a single meal is not controlled by post-absorptive signals. Throughout the animal kingdom, the satiation process appears to be under tight, pre-absorptive sensory control (Smith and Gibbs, 1979). Most animals eat discrete meals which end before absorption of all the ingested nutrients can take place, suggesting that the origin of satiety signals is the gastrointestinal tract (Houpt, 1982). Meals are terminated in expectation of the post-absorptive consequences rather than being solely the result of them, and thus satiety is a state partly specified by the stimulus conditions (Booth, 1985). For example, the sense of taste plays a role in the termination of feeding (Swithers and Hall, 1994). The complexity of these responses emphasizes why voluntary intake is so difficult to predict from first principles: it is, ultimately, a psychological phenomenon. It involves the neural integration of many signals, and is subject not only to the interplay of positive and negative physical and metabolic signals, but also to psychological phenomena such as perceptual constraints and learning (Provenza, 1995). The elucidation of how the integration of signals is affected by the animal’s physiological and mental state remains an important challenge. Attempting a truly mechanistic model of the regulation of food intake (combining taste, pre-absorptive sensation, hormonal responses, conditioning) would present problems of overwhelming complexity, even if all the relevant parameters could be evaluated. A much more parsimonious approach would be to model the functional aspects of the system by addressing the functions that short-term regulation of nutrient intake has evolved to perform.

Prediction of intake and digestion using digesta kinetics models The basic structure of models of intake and digestion is an elaboration of that first proposed by Waldo et al. (1972) and Mertens and Ely (1979), and is illustrated in Fig. 2.1. Clearance of a food constituent from the site of fermentation is the outcome of the competing processes of digestion and passage (which are usually assumed to be firstorder), with respective rates kd and kp, and the fraction digested, D, is therefore a simple function of the relative rates: D = kd/(kd + kp ). Allen and Mertens (1988) and

Mathematical Models of Food Intake and Metabolism in Ruminants

25

Fig. 2.1. Flow diagram of a compartment model sub-unit describing the kinetics of passage and digestion of a substrate (S) as it progresses through a lag phase, undergoes passage and fermentation in the rumen and hindgut. Fermentation causes the substrate to be converted to microbial biomass (Micr) and fermentation products, absorbed from the rumen, small intestine and hindgut. kp and kd are first-order rate constants for passage and digestion.

Mertens (1993) provided comprehensive reviews of elaborations of this basic model. The central assumption of digesta kinetics models is that intake is determined by the physical capacity of the rumen and that the clearance of digesta from the rumen allows further intake to occur. Illius and Allen (1994) made a detailed comparison of the structure and assumptions of published models, which differ largely in their fractionation of food and descriptions of digestion and passage kinetics.

Substrate fractionation Most recent models divide foods into compartments for cell contents, digestible cell wall and indigestible cell wall, in recognition of their different rates of digestion and passage, and assign first-order rate constants to these processes. Compartments should be defined as subsets of the whole that have homogeneous kinetic properties, and these may not necessarily correspond with any physically or chemically definable compartment in the real system. Further disaggregation of cell wall by particle size is usually thought necessary to account for selective retention in the rumen, and possibly different digestion rate, of large particles. The heterogeneity of potentially digestible cell wall

26

A.W. Illius et al.

as a fermentation substrate was established by Van Milgen et al. (1993). They showed differences in the potentially digestible fraction, discrete lag time and digestion rates between cellulose and hemicellulose in both lucerne and wheat straw, with some additional differences in these parameters between particles of different size. Representation of the carbohydrate, protein and lipid components of cell contents may be required to interface with microbial sub-models specified in biochemical terms (Gordon and Illius, 1996). In essence, foods need to be described down to the level of detail that is consistent both with differences in digestion or passage rate or effects on metabolism of sufficient magnitude to affect the system’s dynamics. However, the degree to which information on food composition and component dynamics is available must also be considered. Models that demand a high degree of specification of substrates are limited in their application, due to a shortage of information about food composition and the expense of routinely analysing foods for a wide range of components. Evaluation of highly specified models is also hampered by the limited number of empirical observations conducted at a commensurate level of detail, i.e. of the supply of nutrients to the animal.

Digestion kinetics From a methodological standpoint, ruminal digestion may be regarded as two related processes: the disappearance of feed constituents (as contributors to the dry matter) due to solubilization and microbial fermentation, and the utilization of feed substrates by the microbial population. Studies of digestion kinetics in situ and in vitro typically measure substrate disappearance, which is the natural starting point for the digestion component in digesta kinetics models. In principle, a detailed assessment of digestion kinetic parameters for multiple substrates from in situ studies could be applied to the prediction of digestibility in an appropriate model. In practice, methodological issues and the costs of estimating separate parameters for fractions of each feed present problems for this approach. The many sources of error in estimating lag duration and the rate and potential extent of digestion by in situ techniques have been summarized by Nocek (1988) and others. Artefacts arise due to effects of, for example, pH in the bags, efflux of finely ground material and unduly long lag times of unmasticated forage samples. Not accounting for and estimating the lag time and degradation constant simultaneously can reduce estimated degradation rates markedly, especially if the first bags are removed from the rumen during the lag phase (Dhanoa, 1988). Methodological limitations such as these led Firkins et al. (1998) to stress that more work needs to be done to improve accuracy of estimation of kinetic parameters if models are to predict digestion properly. Until then, uncertainties remain about translating digestion kinetics studied in situ into events occurring in vivo.

Animal factors Most digesta kinetics models have addressed a particular animal type and have not attempted to describe animal effects due to body size and physiological state. The size-

Mathematical Models of Food Intake and Metabolism in Ruminants

27

related effects of species differences can, broadly, be described by allometric relationships, due to the striking association between physiological processes and animal mass (Taylor and Murray, 1987). The duration of physiological events is longer in large animals, and is expected to scale as M0.27. Examples given in Table 2.1 show that mean particulate retention time and the time to comminute large particles scale with exponents not significantly different from 0.27, in the manner of other temporal variables such as the time between successive heart beats or intestinal contractions (Clark, 1927). It will be noted that, despite the high proportion of variance explained (r 2), there is still appreciable prediction error, as indicated by residual coefficient of variation (cv) in the range 13–25% (Table 2.1). Illius and Gordon (1991) showed that the scaling approach works well for interspecific differences. It has a less clear theoretical basis and is less likely to be successful with intraspecific variation (within a group of dairy cows). Animal effects due to physiological state, age and environment are much harder to account for than effects due to size. Chilibroste et al. (1997) adjusted animal effects for physiological state (pregnancy and lactation) according to the Agricultural Research Council (1980). Physiological effects are potentially a major source of error in model predictions, and are of considerable economic importance. Attention needs to be directed to isolating these sources of between-animal variation and modelling the causal variables.

Passage kinetics The values for passage rate that are used in models may be: derived from observation of the food in question; generalized from a range of empirical observations (Sniffen et al., 1992); derived from an allometric function of body mass (Illius and Gordon, 1991); or be a set value for all animals and forages (Mertens and Ely, 1979). Empirical estimates of passage rate are the norm in digesta kinetics models, because mechanistic modelling of passage rate has seldom been attempted. Sauvant et al. (1996) linked particulate outflow to a number of functions such as chewing during rumination and reticular activity, but the approach is still at an investigative stage and the model has not yet

Table 2.1. Allometric relationships between physiological variables and body mass. Range in M (kg)

n

r2

Residual cv (%)

Rumen digesta load (kg DM)

Y=

0.01W1.15(±0.039)

5.5–725

21

0.98

25

Gordon and Illius (1994)

Whole gut mean retention time (h)

Y = 16.7M0.23(±0.026)

1.4–907

48

0.89

13.3

Update of Illius and Gordon (1991)

Retention time of large particles (h)

Y = 7.2N0.69(±0.10)M0.22(±0.058)

40

0.71a

25

Illius and Gordon (1991)

Variable

Allometric expression

Reference

W is body mass less digesta fresh weight; M is body mass; N is indigestible neutral detergent fibre (NDF) concentration; DM is dry matter. aAfter accounting for variation between experiments.

28

A.W. Illius et al.

been tested on many forages. Better information on the pattern of outflow and its relation to chewing and gut motility is needed. The assumption of first-order particle kinetics, when fluxes are a constant proportion of a homogeneous compartment, is increasingly under question. A high flow rate of all nutrients would be expected in the duodenum immediately after feeding if homogeneous pools of specific nutrients are assumed to exist. However, food components arriving in the rumen are not immediately mixed with their respective ruminal compartments and available for digestion and passage. This, together with the rumen’s sacculated structure, the stratification of particulate matter and the effects of particle buoyancy suggest that any food component, such as small particulate matter, may reside in a number of sub-compartments. Therefore, considerable sub-compartmentation may be required to satisfy the requirement for homogeneity. Various mathematical approaches have been used to differentiate between components which are immediately available for passage and those which require hydration, rumination, digestion, etc. Even if a homogeneous pool can be identified close to the reticulo-omasal orifice, constant fractional outflow is only likely to occur if rumen contractions (strength and frequency) are relatively constant throughout the day. Thiago et al. (1992) identified increased numbers of contractions in the first 5 h after consumption of large meals by steers, but total myoelectric activity peaked later in the feeding cycle. The consequence of this later peak was apparent when fractional outflow rates of neutral detergent fibre (NDF) were calculated from data obtained by emptying the rumen at different times after feeding (Gill, 1990). Fractional outflow rate varied between 0.0140 h- 1 (2–6 h) and 0.0410 h- 1 (20–23 h after feeding). P.H. Robinson and M. Gill (unpublished) went a step further and calculated the fractional outflow rates of a number of nutrients during different time periods. The weighted mean fractional rate for NDF was the lowest (0.0242 h- 1) and that for crude protein the highest (0.0825 h- 1). Little diurnal variation was observed for NDF fractional rates, while that for crude protein declined sharply from over 0.115 h- 1 to less than 0.055 h- 1 in immediate response to feeding a protein meal. The passage of heterogeneous digesta was modelled by Matis et al. (1989) as a probabilistic process, using gamma-distributed residence times. They developed a model of passage of heterogeneous digesta with, essentially, time-delay or mixing compartments followed by a final homogeneous compartment from which first-order outflow occurs. Models of this form agree closely with faecal marker flow patterns, and suggest the existence of two particulate pools: a mixing pool that represents a time delay of about 10 h in dairy cows followed by a homogeneous first-order outflow pool. This view is supported by the much closer agreement between observed cell wall digestibility and that predicted by using a two-compartment model with timedependence in the first compartment than by using a model with a single compartment (Huhtanen and Vanhatalo, 1997). Phenomena explaining such age dependency in particulate passage are comminution and buoyancy. Poppi et al. (1981) demonstrated retention times of large particles of tropical forages of c. 11 h and 18 h in sheep and cattle, respectively. Studies with plastic particles indicate that buoyancy also affects passage in both sheep and cattle. Particles with high buoyancy have a lower fractional rate of passage than those with lower buoyancy (Campling and Freer, 1962; Kaske and Engelhardt, 1990). Changes in buoyancy over the time-course of particle digestion and comminution would result in

Mathematical Models of Food Intake and Metabolism in Ruminants

29

passage of particles deviating from first-order kinetics. Such deviation could possibly explain why the model of Illius and Gordon (1991) predicted intake but underestimated digestibility of low-quality forages (Allen, 1996). To examine the possible effects of particle buoyancy, Jessop and Illius (1999) modified the model of Illius and Gordon (1991) to include separate sets of compartments for components of each meal, thereby allowing passage rate to vary with particle age. The rate of passage of small particles, kp¢ , was a function of the proportion of indigestible neutral detergent fibre (INDF), Q, remaining in each particle pool (Allen, 1996). The baseline passage rate, kp, was varied over range R, with threshold INDF concentration Qd and rate s: Ê Ê 1 kp¢ = kp Á 1 + R Á s Ë 1 + e (Q Ë

Qd )

-

1ˆ ˆ ˜ 2 ¯ ˜¯

The parameters R and s were given values of 1.5 and 10 respectively, based on the data of Campling and Freer (1962). Predicted intakes and digestibilities were found to be only slightly sensitive to Qd and rather insensitive to R and s. The optimal value of Qd, determined as the value which gave the best agreement between observed and predicted intake and digestibility, was found not to be constant across foods but to vary consistently with both forage INDF and plant part (leaf or stem). This suggests that chemical and physical factors determine the time-course of buoyancy in a more complex manner than suggested by Allen (1996). The buoyancy model increased predicted digestibility and reduced intake on a forage with low digestion rate (0.02 h- 1) but increased both digestibility and intake rate at a higher digestion rate (0.08 h- 1). The predicted marker excretion pattern was more realistic than from a model omitting the time-delay pool.

Model performance and sensitivity Illius and Allen (1994) reviewed the performance of digesta kinetics models. Few models have been used to predict intake over a wide range of circumstances. The model of Mertens and Ely (1979) under-predicted low dry matter (DM) intakes, over-predicted high DM intakes, and explained only 26% of the variation in the intake observed in the 166 forages tested. The poor relationship between predicted and observed intakes independent of bias is probably due to the use of the same constants for fractional rates of passage and particle size reduction for all forages and animals. The amount of variation in intake explained by the model is less than that observed by using dietary concentrations of fibre (NDF, acid detergent fibre) alone (Mertens and Ely, 1982). Hyer et al. (1991), using a reference data set of 42 forages from the literature, found that intake was systematically under-predicted for low digestibility forages and over-predicted for high digestibility forages. Both Mertens and Ely (1982) and Hyer et al. (1991) suggested that energy demand and not ruminal capacity limits intake of highly digestible forages. That being so, good agreement of prediction with observation from a digesta kinetics model would hardly be expected for high digestibility forages. Illius and Gordon (1991) used data from low digestibility forages and found predicted intake agreed well with observed values, without bias. The variation in intake explained by the model was greater than that predicted by NDF alone (r 2 = 0.10, not significant) or

30

A.W. Illius et al.

INDF (r 2 = 0.18). Chilibroste et al. (1997) used data on digesta and passage kinetics from a variety of sources in the literature and reported close agreement between predicted and observed intake (r 2 = 0.92). Their model attempts to set digesta load as a function of physiological state, to adjust passage for rumen fill, and to model associative effects on feed degradation. Forage digestibility predicted by digesta kinetics models is generally within 15% of observed values (Illius and Allen, 1994), with r 2 in the range 0.5–0.7. The bias in digestibility estimates that frequently occur with these models is more likely to result from poor parameter estimation of digestion kinetics of forages than from inaccurate passage rate or from fundamental flaws in model structure (Illius and Gordon, 1991). The sensitivity of model performance to variation in parameter values is seldom reported, but is of interest as to how accurate models need to be. If model output is sensitive to a particular parameter, then more attention needs to be paid to estimating it accurately than to parameters making little difference to output. The model of Illius and Gordon (1991) showed appreciable sensitivity of intake prediction only to digesta load, particulate passage rate and the proportion of the cell wall that is digestible. Parameters such as digestion rate, lag time, microbial growth efficiency and particle comminution rate caused output to vary by only about 10% of a variation in parameter value. Digestibility predictions were sensitive to the proportion of the cell wall that is digestible, but to little else.

Causal role of digesta load and passage rate in digesta kinetics models In modelling, as in any other form of scientific investigation, progress depends on correctly identifying causal factors. Digesta kinetics models of intake commonly assume that intake is limited by the physical capacity of the rumen and determined by the clearance rate of ruminal digesta, which is dependent on the processes of digestion, particle breakdown and passage rate. It is not necessary to assume that the rumen is ever completely full, merely that some set point of digesta load, normally expressed as mass of DM, serves to regulate intake. Provided that this set-point can be specified, and that digesta turnover is accurately modelled, the amount of intake required to return to the set-point can be predicted. Thus, Illius and Gordon (1991) used a model with a digesta load set-point to achieve a 1:1 relationship (r 2 = 0.61, n = 25, residual cv = 14.5%) between predicted and observed intake (expressed on a metabolic weight basis) in cattle and sheep fed temperate and tropical forages ranging in NDF concentration from 622 to 875 g kg- 1. They derived the digesta load set-point from an allometric analysis (Table 2.1). Despite the apparent success of digesta kinetics models at predicting intake, there are doubts about the validity of the underlying assumptions. Mathison et al. (1995) argued that digesta passage is not merely a property of foodstuffs but is also a function of the propulsive activities of the forestomach, with the implication, at least, that the animal may exert some control over passage. Doubts about the role of digesta load in regulating intake of poor-quality forages have been raised by Weston (1996) and Pitroff and Kothman (1999), arguing that digesta load is not normally at a fixed upper limit, that it varies with the animal’s physiological state and environment, and increases with energy deficit. Over a range of forage quality, Weston (1996) observed a negative cor-

Mathematical Models of Food Intake and Metabolism in Ruminants

31

relation between net energy (NE) intake and digesta load. However, the forages affording the lowest NE intake still allowed the lambs to meet their maintenance requirements for energy at a digesta load of 26% of live mass, suggesting that forages were insufficiently poor to really test for a constant upper limit to digesta load. The best forages were associated with the digesta loads of 11% of live mass, and provided 2.25 times the maintenance energy intake – when energy-based regulation would be expected. Weston (1996) accepted that digesta load must have some upper limit that constrains intake, but argued that metabolic factors such as energy status modify the digesta load the animal will tolerate. Digesta kinetics models would need to incorporate such flexibility in the digesta load constraint to account for variations in intake due to the animal’s physiological status. It is harder to address the question of whether passage rate is the driving variable in digesta residence in the fermentation compartment, or whether the animal’s demand for nutrients drives both intake rate and passage rate. Illius and Gordon (1991) showed that retention time scales approximately as expected, with M0.27. But since digesta load scales as about M1 and energy and protein requirements scale as M0.73, it could equally be argued that retention time must scale as M1/M0.73, or M0.27 when animals are eating to requirement. The argument that other temporal variables (see above) scale as M0.27 does not really resolve this debate about which is the fundamental causal variable because the reason why metabolic rates and time scale allometrically as they do is unclear. Given the frequently observed negative correlation between NDF concentration and intake (Jung and Allen, 1995), which implies that something other than energy demand constrains intake, we are still inclined to the view that retention time is not largely under the control of the animal, and that constraint to passage rate induced by feed characteristics is a causal factor in the intake of low-quality forages. In summary, the current generation of digesta kinetics models can apparently predict intake of low-to-medium quality forages by animals with modest nutrient requirements, but deeper knowledge of the mechanisms underlying control of digesta load and passage rate are needed to allow such models to predict how intake varies with physiological state and nutrient demand.

Metabolic models Microbial metabolism The objective of metabolic models that focus on substrate metabolism by microbes is the prediction of nutrient supply to the animal (Black et al., 1980–1981; Baldwin et al., 1987b; Dijkstra et al., 1992, 1998). In the earliest model, substrate supply to a single microbial pool followed first-order kinetics (Black et al., 1980–1981). The energy (ATP) released from fermentation of substrate is used to meet the microbial maintenance requirement and in excess of this is used to drive microbial growth. Growth can be limited by either ATP yield or fermentable nitrogen supply. In the models of Baldwin et al. (1987b) and Dijkstra et al. (1992, 1998), supply of substrate is a secondorder process, being influenced by both substrate concentration and microbial mass following Michaelis–Menten kinetics, but for which the derivation of appropriate parameter

32

A.W. Illius et al.

values is problematic. A single pool of microbes is divided between large particles, small particles and water in the model of Baldwin et al. (1987b), and therefore microbes flow out of the rumen at a composite rate. The models of Dijkstra et al. (1992, 1998) describe three microbial pools representing amylolytic and fibrolytic microorganisms and protozoa. Microbial maintenance requirements vary between these pools, as do substrate specificities. Amylolytic and protozoal pools had variable composition such that starch levels could increase at high nutrient availability and provide a reserve of carbohydrate at times of nutrient shortage. Protozoa engulf amylolytic and fibrolytic microorganisms in relation to their relative pool sizes although engulfment rate was reduced as protozoal starch content increased. The outflow rates of each pool differ, amylolytic microorganisms being washed out of the rumen at the liquid passage rate, fibrolytic ones at the solid passage rate and protozoa at a lower rate. None of the models represent effects of low ruminal pH on cellulolytic activity although such effects are being studied (Sauvant, 1998). These models therefore place less emphasis on digesta kinetics. However, the shift in focus from digesta disappearance to the metabolism of feed constituents does not liberate such models from the exigencies of properly defining the digestion kinetics of those constituents. These constituents are the substrates that are inputs to microbial models, which are usually modelled at a given steady-state rate of intake. Bannink et al. (1997a) compared three sophisticated models of rumen function (Baldwin et al., 1987b; Danfær, 1990; Dijkstra et al., 1992) using inputs from seven diets with very complete observations available of rumen dynamics, and found that the models gave markedly different results. Much of the reason for this stems from the use of some concepts and parameter inputs, for instance concerning particle dynamics and the partitioning of carbohydrate fermentation, that can not be estimated from rumen observations. Although the model of Dijkstra et al. (1992) uses digestion kinetics parameters derived from in situ studies, the other two do not, relying instead on the microbial utilization of substrate to drive digestion. Bannink et al. (1997a) concluded that, without any input from in situ studies, microbial metabolism models are unlikely to be able to replicate observed differences in feed degradation. Explanations of the divergence between model predictions of volatile fatty acids (VFA) production and values measured in vivo, may relate both to inadequacies in modelling fermentation stoichiometry and absorption kinetics (Bannink et al., 1997b) and to artefacts in estimation of VFA production (Beever, 1993). In general, rumen models are highly sensitive to changes in passage rates, because of their effects on the time for which substrates are accessible by the microbes in the rumen. However, the study of passage rates has received much less attention than that of digestion rates and thus it is difficult to assess the degree of error introduced. For example, over a meal cycle it has been shown that there is marked variation in the liquid outflow rate from the rumen (Warner and Stacy, 1968). Depending on the characteristics of the feed this can have quite different effects on the pattern of nutrient absorption. There are few data describing the patterns of liquid and particulate outflow from the rumen during a meal cycle, i.e. in non-steadystate conditions. Nearly all experimental work has attempted to maintain steady-state by constant, low-level feeding over the 24 h period where there is unlikely to be large fluctuations in rumen dynamics. Considering the large number of parameter values required as inputs by sophisticated models of rumen function, they have limited application for predictive purposes

Mathematical Models of Food Intake and Metabolism in Ruminants

33

at present, but are valuable research tools which need to be developed further (Bannink et al., 1997a).

Models of animal response to nutrients Such models use, explicitly or implicitly, the nutrients produced from fermentation and digestion of ingested food as inputs. They aim to predict the consequences in terms of change in state of the animal (gain or loss of body mass) and synthesis of products such as milk. The models differ widely in the level of detail used, ranging from those that operate using metabolizable energy (ME) and protein (MP, e.g. AFRC, 1993) to those that require a detailed description of the profile of absorbed nutrients (Baldwin et al., 1987a). The ME and MP system provides an example of an aggregated, phenomenological approach in which ME and MP are used following a set of rules determined from empirical observations. Although distinction is drawn between protein and energy supply, ME includes the energy content of absorbed amino acids regardless of whether the amino acids will be catabolized or the efficiency with which energy-yielding metabolites will be used. The empirical relationships on which the model is based vary depending on mature size and sex of the animal to account for such differences in maintenance requirements, nutrient partitioning and composition of gain. A simple and elegant model of nutrient use has recently been published by Emmans (1997) that has general applicability to a wide range of animal species. This approach has been used as the basis of predicting requirements and intake (Oldham, 1999) and the description of nutrient supply and requirement can be at a relatively aggregated (protein, and non-protein energy) or more detailed (individual amino acids, carbohydrate and lipid) level depending on the information available. The most detailed models produced are based on a description of the major metabolic pathways (e.g. Baldwin et al., 1980, 1987a; Gill et al., 1984; Illius and Jessop, 1995). These models consist of a set of metabolic steps that vary from describing a single enzymic reaction to a complete metabolic pathway. Metabolites are state variables and their oxidation or conversion into products (such as tissue protein or milk protein) is determined by the rates of the metabolic pathways involved. Michaelis–Menten kinetics are used such that the overall flux through a pathway depends on the potential activity of that pathway, Vmax, the Michaelis–Menten constant, km, and the concentration of the substrate. Additional terms representing end product inhibition or sigmoidal responses to substrate or hormonal effects on Vmax and km may be added. Such models are parameterized (i.e. the values of Vmax and km derived) from experiments in vitro performed on tissue slices or isolated cells. The level of aggregation varies between models. Gill et al. (1984) treated all transactions as occurring in a single pool. Baldwin et al. (1987a) separated metabolism between viscera, adipose tissue and the remaining lean body and allow perfect communication between them. Illius and Jessop (1995) divided metabolism between viscera and the rest of the body with transfer of nutrients determined by rates of blood flow and diffusion through extracellular water. The number of parameters required for such models is large and it is unclear as to how these might vary between species or even within species across different physiological states or degrees of maturity.

34

A.W. Illius et al.

The main strengths of metabolic modelling are in testing where knowledge is inadequate, and in the coordination of experimental work. One example of synergistic links between modelling and an experimental programme was research aimed at improving understanding of voluntary intake by growing ruminants. Initially, a simplified model of growth (Gill et al., 1984) represented all the protein in the body as one pool, although it is known that rates differ between organs. A separate model was then developed (Gill et al., 1989) using experimental data on protein turnover rates generated by work at the tissue level. This second model identified the importance of the contribution of protein metabolism in the liver and gut to overall energy maintenance, which identified the need for research on the metabolism of the liver and gut. These initial studies increased understanding of the differences in metabolism of these tissues in steers fed forage and concentrate, and thus indirectly to an increased understanding of factors controlling intake.

Modelling studies of metabolic constraints on intake There have been few attempts to model the interaction of digestive and metabolic constraints on feed intake, despite the probable importance of this interaction. One of the reasons is the lack of data to parameterize metabolism models at the level of the whole animal. The most detailed models, resulting from research programmes integrating in vitro studies of tissues with a modelling exercise at the whole animal level, still require that intake be specified. The first attempt at integrating concepts of metabolic control and physical limitation of eating behaviour and feed intake in ruminants was made by Forbes (1980). The model assumed that animals would satisfy their energy requirements, subject to a physical constraint imposed by rumen capacity. Literature estimates of the relation between food digestion and passage rates were used to derive estimates of the rates of energy absorption and the quantity of food residues. Meal patterning arose from the comparison of the rate of energy supplied by digestion with the rate of energy utilization in maintenance, lactation and growth; thresholds being assigned to initiate and terminate eating, subject to physical constraints. The model was not particularly sensitive to the values used for these thresholds. An alternative mechanism to that of dual thresholds is that of positive feedback, in which positive reinforcement of eating increases the motivation to eat and temporarily overcomes negative stimuli arising from nutrient absorption (Houston and Sumida, 1985). Since the upper limit on energy utilization and the lower limit imposed by gut capacity and passage rate are inputs to the Forbes model, thereby placing bounds on food intake, intake is predicted well by the model. The pattern of meals was also realistic, suggesting that the model gave a good representation of the surface phenomena, and the model clearly had heuristic value in showing that ruminant feeding behaviour could be simulated numerically. The model of Fisher et al. (1987) was also used to study the integration of chemostatic and physical constraints on ruminant feed intake, one of the objectives being to redress the omission of chemostatic feedback from the model of Mertens and Ely (1979). The method of incorporating feedback controls was to strike a balance between negative and positive stimuli, rather than there being separate set-points or thresholds for each. They achieved this by a mathematical formulation expressing

Mathematical Models of Food Intake and Metabolism in Ruminants

35

intake as a function of rumen distension and nutrient flow. The function contains a double exponential term intended to relate the strength of each stimulus in relation to the other (i.e. chemostasis being generally weaker than distension signals, especially at high distension, but stronger at low distension). Investigation of the effect of varying the parameter values demonstrated how the relative strength of the stimuli could be altered, and values were found which gave good approximations between observed and predicted intakes. This model requires the use of some imaginary parameters, and thus can be said to be exploratory. The same applies to the model of Sauvant et al. (1996) insofar as it represents feeding motivation in a ‘feeding decision submodel’ which includes energy status and food palatability. The model determines eating and rumination behaviour to predict intake and passage in an innovative manner, based around more conventional rumen and microbial submodels. The model incorporates a function to relate feeding motivation to fill, rather than using an upper limit to fill as a fixed constraint, but requires careful parameterization, particularly for assumed effects of palatability. Poppi et al. (1994) used the model of animal metabolism developed by Gill et al. (1984) to examine the integration of intake regulation, the approach being to identify pathways that could limit intake and to calculate the first limiting pathway or factor. Both physical and metabolic pathways were examined: namely, instantaneous intake rate, faecal output, rumen fill, genetic potential for protein deposition, heat dissipation and ATP degradation. The ‘ATP degrader’ required by their model to avoid a build-up of ATP was used as an indicator of excessive energy intake or nutrient imbalance. Modifications of the model by Illius and Jessop (1995) removed the requirement for an ATP degrader, instead of which a build-up of acetate indicated excessive energy intake or inadequate glycerol precursors for acetate clearance. Poppi et al. (1994) found energy excess to be limiting, sometimes simultaneously with other pathways, on diets ranging from poor-quality forages to cereals, and the authors concluded that this is indicative of the wide range of dietary conditions under which nutrient balance is implicated in intake regulation.

Future challenges to modelling intake and metabolism Optimizing the quantity, ratio and diurnal pattern of nutrients absorbed by ruminants remains an important goal for nutritional and modelling research. One clear requirement is further integration of digesta kinetics and metabolic models for the purpose of incorporating both animal and diet effects on intake, nutrient supply and utilization. This will further require a suitable framework for modelling the effects of metabolic signals on eating and digesta kinetics. It is likely that functional approaches will be more successful than wholly mechanistic ones, given the complexity of integrating psychological, biochemical and physical signals. Combining empirical estimates of feed degradation rates with models of microbial metabolism acknowledges the effects of physical feed characteristics, but such models still require improvements in the understanding and parameterization of fermentation pathways if VFA production is to be predicted more accurately. Models that describe microbial or animal metabolism in terms of known biochemical pathways generally use parameter values estimated to yield sensible results in particular

36

A.W. Illius et al.

circumstances. The clear problem facing these very detailed models is the difficulty of reaching satisfactory estimates of critical parameter values, especially when experimental work and modelling are not carried out in harmony (Baldwin and Sainz, 1995). Even when they are (Gill and Beever, 1991), it is not clear that single, static values for parameters is a satisfactory representation of dynamic and adaptive metabolic systems. Metabolic modelling also needs to get away from steady-state conditions, in order to accord with the reality of variable inputs and the many time-delaying components of the system (Sauvant and Van Milgen, 1995). Further work is required on the mechanistic modelling of passage, and on the significance of variable fractional outflow rates. Where daily nutrient supply is the output of interest, the effect of variable outflow will only be important for nutrients with very short mean retention times, which are also digestible, e.g. highly degradable proteins. For models which include representation of hormonal effects dependent on nutrient absorption, hour-by-hour changes in the flow of nutrient may be of considerable significance, but flows alone are inappropriate, since further delays or changes in profiles may occur during absorption through the intestinal wall. The consequences of variable flow rates on these different processes need to be examined.

References AFRC (1993) Energy and Protein Requirements of Ruminants. CAB International, Wallingford, UK. Agricultural Research Council (1980) The Nutritional Requirements of Ruminant Livestock. Commonwealth Agricultural Bureaux, Slough. Allen, M.S. (1996) Physical constraints on voluntary intake of forages by ruminants. Journal of Animal Science 74, 3063–3075. Allen, M.S. and Mertens, D.R. (1988) Evaluating constraints on fiber digestion by rumen microbes. Journal of Nutrition 118, 261–270. Baldwin, R.L. and Sainz, R.D. (1995) Bottlenecks for improvement of future models in animal nutrition science. In: Danfaer, A. and Lescoat, P. (eds) Proceedings of IV International Workshop on Modelling Nutrient Utilization in Farm Animals. National Institute for Animal Science, Jele, Denmark, pp. 163–171. Baldwin, R.L., Lucas, H.L. and Cabrera, R. (1970) Energetic relationships in the formation and utilization of fermentation end-products. In: Phillipson, A.T. (ed.) Physiology of Digestion and Metabolism in the Ruminant. Oriel Press Ltd, Newcastle upon Tyne, UK, pp. 319–334. Baldwin, R.L., Smith, N.E., Taylor, J. and Sharp, M. (1980) Manipulating metabolic parameters to improve growth rate and milk secretion. Journal of Animal Science 51, 1416. Baldwin, R.L., France, J. and Gill, M. (1987a) Metabolism of the lactating cow. I. Animal elements of a mechanistic model. Journal of Dairy Research 54, 77–105. Baldwin, R.L., Thornley, J.H.M. and Beever, D.E. (1987b) Metabolism of the lactating cow. II. Digestive elements of a mechanistic model. Journal of Dairy Research 54, 107–131. Bannink, A., DeVisser, H., Dijkstra, J. and France, J. (1997a) Impact of diet-specific input parameters on simulated rumen function. Journal of Theoretical Biology 184, 371–384. Bannink, A., DeVisser, H., Klop, A., Dijkstra, J. and France, J. (1997b) Causes of inaccurate prediction of volatile fatty acids by simulation models of rumen function in lactating cows. Journal of Theoretical Biology 189, 353–366. Beever, D.E. (1993) Rumen function. In: Forbes, J.M. and France, J. (eds) Quantitative Aspects of Ruminant Digestion and Metabolism. CAB International, Wallingford, pp. 187–215. Black, J.L., Beever, D.E., Faichney, G.J., Howarth, B.R. and Graham, N.McC. (1980–81)

Mathematical Models of Food Intake and Metabolism in Ruminants

37

Simulation of the effects of rumen function on the flow of nutrients from the stomach of sheep: part 1 – Description of a computer program. Agricultural Systems 6, 195–219. Booth, D.A. (1985) Food-conditioned eating preferences and aversions with interoceptive elements: Conditioned appetites and satieties. Annals of the New York Academy of Science 443, 22. Campling, R.C. and Freer, M. (1962) The effect of specific gravity and size on the mean time of retention of inert particles in the alimentary tract of the cow. British Journal of Nutrition 16, 507–518. Chilibroste, P., Aguilar, C. and García, F. (1997) Nutritional evaluation of diets. Simulation model of digestion and passage of nutrients through the rumen-reticulum. Animal Food Science and Technology 68, 259–275. Clark, A.J. (1927) Comparative Physiology of the Heart. Cambridge University Press, Cambridge, UK. Danfær, A. (1990) A dynamic model of nutrient digestion and metabolism in lactating dairy cows. PhD Thesis, Report 671, National Institute of Animal Science, Denmark. Dhanoa, M.S. (1988) On the analysis of dacron bag data for low degradability feeds (Research Note). Grass and Forage Science 43, 441–444. Dijkstra, J., Neal, H.D.St.C., Beever, D.E. and France, J. (1992) Simulation of nutrient digestion, absorption and outflow in the rumen: model description. Journal of Nutrition 122, 2239–2256. Dijkstra, J., France, J. and Tamminga, S. (1998) Quantification of the recycling of microbial nitrogen in the rumen using a mechanistic model of rumen fermentation processes. Journal of Agricultural Science, Cambridge, 130, 81–94. Egan, A.R. (1977) Nutritional status and intake regulation in sheep. VIII Relationships between the voluntary intake of herbage by sheep and the protein/energy ratio in the digestion products. Australian Journal of Agricultural Research 28, 907. Emmans, G.C. (1997) A method to predict the food intake of domestic animals from birth to maturity as a function of time. Journal of Theoretical Biology 186, 189–199. Faichney, G. (1993) Digesta flow. In: Forbes, J.M. and France, J. (eds) Quantitative Aspects of Ruminant Digestion and Metabolism. CAB International, Wallingford, pp. 53–86. Firkins, J.L., Allen, M.S., Oldick, B.S. and St-Pierre, N.R. (1998) Modeling ruminal digestibility of carbohydrates and microbial protein flow to the duodenum. Journal of Dairy Science 81, 3350–3369. Fisher, D.S., Burns, J.C. and Pond, K.R. (1987) Modeling ad libitum DM intake by ruminants as regulated by distension and chemostatic feedbacks. Journal of Theoretical Biology 126, 407. Forbes, J.M. (1980) A model of the short-term control of feeding in the ruminant: effects of changing animal or feed characteristics. Appetite 1, 21. Forbes, J.M. and France, J. (1993) Quantitative Aspects of Ruminant Digestion and Metabolism. CAB International, Wallingford, pp. 1–10. Gill, M. (1990) Future for whole-animal metabolism models. In: Robson, A.B. and Poppi, D.P. (eds) Proceedings of Third International Workshop on Modelling Digestion and Metabolism in Farm Animals. Lincoln University, New Zealand, pp. 333–345. Gill, M. and Beever, D.E. (1991) Modelling as an aid to nutrition research. In: Isotope and Related Techniques in Animal Production and Health. Proceedings of an International Symposium on Nuclear and Related Techniques in Animal Production and Health jointly organised by IAEA and FAO, Vienna, April 1991. Vienna: IAEA, pp. 171–182. Gill, M., Thornley, J.H.M., Black, J.M., Oldham, J.D. and Beever, D.E. (1984) Simulation of the metabolism of absorbed energy-yielding nutrients by the young sheep. British Journal of Nutrition 52, 621. Gill, M., France, J., Summers, M., McBride, B.W. and Milligan, L.P. (1989) Mathematical integration of protein metabolism in growing lambs. Journal of Nutrition 119, 1269–1286.

38

A.W. Illius et al. Gordon, I.J. and Illius, A.W. (1994) The functional significance of the browser-grazer dichotomy in African ruminants. Oecologia 98, 167–175. Gordon, I.J. and Illius, A.W. (1996) The nutritional ecology of ruminants: a reinterpretation. Journal of Animal Ecology 65, 18–28. Houpt, K.A. (1982) Gastrointestinal factors in hunger and satiety. Neuroscience and Biobehavioral Reviews 6, 145. Houston, A. and Sumida, B. (1985) A positive feedback model for switching between two activities. Animal Behaviour 33, 315. Huhtanen, P. and Vanhatalo, A. (1997) Ruminal and total cell-wall digestibility estimated by a combined in situ method utilizing mathematical models. British Journal of Nutrition 78, 583–598. Hyer, J.C., Oltjen, J.W. and Galyean, M.L. (1991) Evaluation of a feed intake model for the grazing beef steer. Journal of Animal Science 69, 836–842. Illius, A.W. and Allen, M.S. (1994) Assessing forage quality using integrated models of intake and digestion by ruminants. In: Fahey, G.C. (ed.) Forage Quality, Evaluation, and Utilization. University of Nebraska, Lincoln, Nebraska, pp. 869–890. Illius, A.W. and Gordon, I.J. (1991) Prediction of intake and digestion in ruminants by a model of rumen kinetics integrating animal size and plant characteristics. Journal of Agricultural Science, Cambridge 116, 145–157. Illius, A.W. and Jessop, N.S. (1995) Modelling metabolic costs of allelochemical ingestion by foraging herbivores. Journal of Chemical Ecology 21, 693–719. Illius, A.W. and Jessop, N.S. (1996) Metabolic constraints on voluntary intake in ruminants. Journal of Animal Science 74, 3052–3062. Jessop, N.S. and Illius, A.W. (1999) Modeling the influence of buoyancy on particle dynamics in the foregut of ruminants. In: Stone, G., Forbes, T.D.A., Stuth, J.W. and Byers, F.M. (eds) Veterinary International Symposium on the Nutrition of Herbivores 10–16 April, San Antonio, Texas (CD-ROM). Jung, H.G. and Allen, M.S. (1995) Characteristics of cell walls affecting intake and digestibility of forages by ruminants. Journal of Animal Science 73, 2774–2790. Kaske, M. and Engelhardt, W. von (1990) The effect of size and density on mean retention time of particles in the gastrointestinal tract of sheep. British Journal of Nutrition 63, 457–465. Laredo, M.A. and Minson, D.J. (1973) The voluntary intake, digestibility, and retention time by sheep of leaf and stem fractions of five grasses. Australian Journal of Agricultural Research 24, 875–888. Mathison, G.W., Okine, E.K., Vaage, A.S., Kaske, M. and Milligan, L.P. (1995) Current understanding of the contribution of the propulsive activities in the forestomach to the flow of digesta. In: Englehardt, W.V., Leonhard-Marek, S., Breves, G. and Gieseck, D. (eds) Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction. Ferdinand Enke Verlag, Stuttgart, p. 23. Matis, J.H., Wehrly, T.E. and Ellis, W.C. (1989) Some generalized stochastic compartment models for digesta flow. Biometrics 45, 703–720. Mertens, D.R. (1993) Rate and extent of digestion. In: Forbes, J.M. and France, J. (eds) Quantitative Aspects of Ruminant Digestion and Metabolism. CAB International, Wallingford, pp. 13–52. Mertens, D.R. and Ely, L.O. (1979) A dynamic model of fiber digestion and passage in the ruminant for evaluating forage quality. Journal of Animal Science 49,1085–1095. Mertens, D.R. and Ely, L.O. (1982) Relationship of rate and extent of digestion to forage utilization – A dynamic model evaluation. Journal of Animal Science 54, 895–906. Nocek, J.E. (1988) In situ and other methods to estimate ruminal protein and energy digestibilty: A review. Journal of Dairy Science 71, 2051–2069. Oldham, J.D. (1999) Genotype ¥ nutrition interactions in herbivores. In: Jung, H.-J.G. and

Mathematical Models of Food Intake and Metabolism in Ruminants

39

Fahey, G.C., Jr (eds) Vth International Symposium on the Nutrition of Herbivores. American Society of Animal Science, Savoy, Illinois, pp. 482–504. Pitroff, W. and Kothman, M.M. (1999) Regulation of intake and diet selection by herbivores. In: Jung, H.-J.G. and Fahey, G.C., Jr (eds) Vth International Symposium on the Nutrition of Herbivores. American Society of Animal Science, Savoy, Illinois, pp. 366–422. Poppi, D.P., Minson, D. and Ternouth, J.H. (1981) Studies of cattle and sheep eating leaf and stem fractions of grasses. III. The retention time in the rumen of large feed particles. Australian Journal of Agricultural Research 32,123–137. Poppi, D.P., Gill, M. and France, J. (1994) Integration of theories of intake regulation in growing ruminants. Journal of Theoretical Biology 167, 129. Provenza, F.D. (1995) Postingestional feedback as an elementary determinant of food preference and intake in ruminants. Journal of Range Management 48, 2. Sauvant, D. (1998) In vitro Techniques for measuring Nutrient Supply to Ruminants. British Society of Animal Science Occasional Publication No. 22. Sauvant, D. and Van Milgen, J. (1995) Dynamic aspects of carbohydrate and protein breakdown and the associated microbial matter synthesis. In: Englehardt, W. von, Leonhard-Marek, S., Breves, G. and Gieseck, D. (eds) Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction. Ferdinand Enke Verlag, Stuttgart, pp. 71–91. Sauvant, D., Baumont, R. and Faverdin, P. (1996) Development of a mechanistic model of intake and chewing activities of sheep. Journal of Animal Science 74, 2785–2802. Sinclair, L.A., Garnsworthy, P.C., Newbold, J.R. and Buttery, P.J. (1993) Effect of synchronizing the rate of dietary energy and nitrogen release on rumen fermentation and microbial protein synthesis in sheep. Journal of Agricultural Science, Cambridge 120, 251. Sniffen, C.J., O’Connor, J.D., Van Soest, P.J., Fox, D.G. and Russell, J.B. (1992) A net carbohydrate and protein system for evaluating cattle diets: II. Carbohydrate and protein availability. Journal of Animal Science 70, 3562–3577. Smith, G.P. and Gibbs, J. (1979) Postprandial satiety. Progress in Psychobiology and Physiological Psychology 8, 179. Swithers, S.E. and Hall, W.G. (1994) Does oral experience terminate ingestion? Appetite 23, 113. Taylor, St.C.S. and Murray, J.I. (1987) Genetic aspects of mammalian survival and growth in relation to body size. In: Hacker, J.B. and Ternouth, J.H. (eds) The Nutrition of Herbivores. Academic Press, Sydney, pp. 487–533. Tolkamp, B.J. and Ketelaars, J.J.M.K. (1992) Toward a new theory of feed intake regulation in ruminants 2. Costs and benefits of feed consumption: an optimization approach. Livestock Production Science 30, 297. Thiago, L.R.L., Gill, M. and Sissons, J.W. (1992) Studies of method of conserving grass herbage and frequency of feeding in cattle. 2. Eating behaviour, rumen motility and rumen fill. British Journal of Nutrition 67, 319–336. Van Milgen, J., Berger, L.L. and Murphy, M.R. (1993) Digestion kinetics of alfafa and wheat straw assuming heterogeneity of the potentially digestible fraction. Journal of Animal Science 71, 1917–1923. Waldo, D.R., Smith, L.W. and Cox, E.L. (1972) Model of cellulose disappearance from the rumen. Journal of Dairy Science 55, 125–129. Warner, A.C.I. and Stacy, B.D. (1968) The fate of water in the rumen 2. Water balances throughout the feeding cycle in sheep. British Journal of Nutrition 22, 389–410. Weston, R.H. (1996) Some aspects of constraints to forage consumption by ruminants. Australian Journal of Agricultural Research 47, 175–197.

3

Control of Salivation and Motility of the Reticulorumen by the Brain in Sheep W.L. GROVUM1 AND J.S. GONZALEZ2 1Department

of Biomedical Sciences, Ontario Veterinary College, University of Guelph, Guelph, Ontario, Canada; 2Department of Animal Production, University of Leon, Leon, Spain

Introduction Brain–gut interactions are critical in salivation, gastric motility and ingestive behaviour in ruminants but the sites in the brain controlling these functions are largely unknown. Better understanding in these areas may help to resolve some of the numerous problems associated with feeding high concentrate diets to grazing ruminants in intensive production systems. Examples of the problems include acidosis, inappetence, vagal indigestion and stasis. The basic mechanisms limiting roughage consumption may also be addressed to enhance the productivity of ruminants when they are at pasture or being fed preserved forages. No papers exist on the sites in the ruminant brain which control salivation, except for Grovum and Gonzalez (1994), whereas the central controls over motility of the reticulorumen are limited to the brain stimulation work of Bell and Lawn (1955), Dussardier (1960) and Andersson et al. (1958) and to the single unit studies of Beghelli et al. (1963), Howard (1970) and Harding and Leek (1971). The bases for sequencing and pacing reticuloruminal contractions are largely unknown. Although it is known that the reticulo-omasal orifice and the reticular groove are under vagal control (Titchen and Newhook, 1975), the origins of orifice constrictor and dilatory neurons and of the groove closure neurons in the brainstem have never been reported. Most of the brain stimulation work reported here has been published recently as short communications (Grovum and Gonzalez, 1994, 1998a,b,c; Grovum, 1998).

Methodology used to identify central sites controlling salivation and gastric motility The simple premise for using electrical stimulation was that the preganglionic cell bodies of facial, glossopharyngeal and vagal neurons were located in discrete regions of the brainstem and that systematic stimulation would eventually excite them, induce © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

41

42

W.L. Grovum and J.S. Gonzalez

saliva secretion and/or contractions of the stomach and hence disclose their location in a three-dimensional map relative to obex. All experiments were acute, the sheep being anaesthetized with sodium pentobarbital. Fluids, consisting of normal saline with 5% glucose and 20 mEq KCl l- 1 added, were given at 1 drop s- 1 (about 3 l per sheep). Normal body temperature was maintained by circulating warm water through a radiator under the operating table. Saliva, dripping from cannulae inserted into the parotid and submandibular ducts via the mouth, was measured with infrared drop counters and their outputs were recorded as vertical deflections on a grass polygraph. Pressures, measured in anaesthetic breathing bags inserted surgically into the reticulum (400 ml inflation) and the ventral sac of the rumen (1400 ml) and in a catheter flushed with heparinized saline and inserted into the saphenous artery, were also recorded. The sheep’s head was secured in a stereotaxic instrument and, after opening the skull and cutting the dura, the brainstem was exposed by removing the cerebellum with suction and ligating the tentorium on the right. The palate-bar was adjusted vertically to make the floor of the IV ventricle horizontal. This was not done when the forebrain was studied because the skull and dura were opened only over the cerebrum. Current spread from the tip of the concentric electrode was found to be less than 1 mm downward in studies on salivation.

Investigations into the central control of saliva secretion Brainstem sites from which parotid and mandibular secretions were evoked The three-dimensional characteristics of the parotid and submandibular centres found in two of six sheep in the work of Grovum and Gonzalez (1994) are illustrated in Fig. 3.1. The parotid centres were located an average of 2 mm caudal to the submandibular centres and otherwise overlapped them in the rostral–caudal plane (see lower portions of each graph extended below the grid representing the base of the brain). In threedimensional coordinates, the combined centres were located on average between 7 and 16 mm rostral to obex and from 1 to 11 mm lateral from the midline (see lower portions again). The centres were 2–5 mm deep (see upper graphs), the dorsal edges being on average 1 mm below the floor of the fourth ventricle near midline and 6 mm below it at the lateral extremes. Here, the excitable parotid tissue extended in most cases down Fig. 3.1. (opposite) The three-dimensional topography of sites (centres) in the brainstem of two sheep where electrical stimulation evoked saliva secretion by the parotid (嘷) and the mandibular (䊱) glands. The grid in the middle of each graph delimits the base of brain. The points below the grid are simply extensions from those above to highlight the coordinates of the centres in the rostral–caudal and lateral planes. These results also demonstrate the variability that exists in the locations of the sites in different sheep. Vertical scale – 0 mm is the floor of the IV ventricle. Rostral–caudal or anterior–posterior (A–P) stereotaxic plane – 0 is obex and positive numbers are rostral to it. Lateral plane – 0 is midline. VII and IX – Facial and glossopharyngeal nerves respectively.

Control of Salivation and Motility of the Reticulorumen

43

44

W.L. Grovum and J.S. Gonzalez

to the base of the brain. From a dorsal perspective (see lower portions), the orientation of most of the centres was essentially in the lateral plane except that the submandibular centres angled slightly cranially from midline. The combined centres in four sheep were located between cranial nerves VII and IX in two, adjacent to VII in one and adjacent or slightly caudal to IX in the other. In the two sheep in which only the submandibular centres were found (not illustrated), one was located between VII and IX and the other was more adjacent to IX than between VII and IX. This variability with respect to the cranial nerves was in addition to that for the stereotaxic coordinates of landmarks in the sheep brain of up to 5 mm in the vertical and anterior–posterior (A–P) planes observed by McKenzie and Smith (1973). The practical significance of such variability for future work is that the secretion sites will have to be localized with stimulation before singleunit recordings are made. The finding of functionally separate sites controlling parotid and submandibular secretion is in agreement with the fact that the parotid gland secretes during resting, eating and rumination, whereas the submandibular gland produces saliva only during eating (Denton, 1957; Carr, 1984). A functional separation of these two central systems existed, even though the neurons controlling these glands were anatomically largely intermingled.

Frontal lobe site where parotid secretion was evoked Stimulating the apex of the frontal cortex with bipolar side-by-side electrodes 15–20 mm from midline and 0–15 mm on the underside of the curvature of the brain elicited secretions from the parotid glands in ten sheep (Grovum and Gonzalez, 1998b). The effect was increased when the strength of stimulation was increased from 8–20 V (Fig. 3.2). Chewing movements, swallowing, and an inhibition of respiration, were frequently but not always associated with the secretory responses described. Forestomach motility was not affected. The normal inputs, which excite or inhibit this ‘parotid’ site, are unknown except for the fact that stimulating the olfactory bulbs (8–20 V with bipolar side-by-side electrodes) where they attached to the frontal lobe had no effect on salivation.

A putative osmosensitive site in the amygdala inhibiting parotid secretion When 2 ml of a 2 Osm solution of glucose, polyethylene glycol (PEG)-200, NaCl, sucrose or urea were injected separately into the jugular vein, or into the circulations of the parotid gland (isolated surgically except for one artery, one vein and nerve supply) and the brain (retrograde injections through the mandibular artery exposed in jaw and cannulated), parotid secretion was decreased on average by 1, 20 and 34% respectively. All site means differed from one another (P < 0.001; Grovum, 1998). The effects in the gland were substrate specific and difficult to understand, and furthermore, were achieved at unphysiological tonicity levels. However, when the solutes were introduced directly into the circulation to the brain, via retrograde injections into the mandibular artery, the depressions in parotid secretion appeared to be mediated by an osmosensitive mechanism, because the inhibition by NaCl (% decrease from control values =

Control of Salivation and Motility of the Reticulorumen

45

Fig. 3.2. Parotid secretion (PS) evoked in four sheep by stimulating the frontal cortex in the sheep brain on average 15–20 mm from midline and either on its apex or down to 15 mm on its underside using 8 and 20 V. The vertical deflections in the traces indicate drops. Secretions by the submandibular gland (SMS) were not affected. The marker line (middle trace) showed when stimuli were applied (marker signal was not activated for sheep 15).

36bc) did not differ from that of glucose (32ab), PEG-200 (35ab), or sucrose (43c) (the means with similar superscripts were not different (P > 0.05)). This effect of NaCl could therefore be accounted for completely by its osmotic characteristics. The fact that the inhibition by urea (29a) was 83% of that for the other solutes was also noteworthy. In fact, the strong responses to both glucose and urea indicated that the inhibition of parotid secretion was mediated by an osmosensitive system different from the osmoreceptors in the organum vasculosum of the lateral terminalis (OVLT) near the III ventricle which evoke thirst and water conservation through antidiuretic hormone (ADH) release. The osmoreceptor terminology is justified by a direct effect of solute on the sensors due to the fact that the OVLT region is devoid of a blood–brain barrier. The reason for claiming that the system inhibiting salivation is different, is that thirst and ADH release are affected either little or not at all by injections of glucose and urea according to Fitzsimons (1989), because they can apparently diffuse into the osmoreceptors and fail to excite them as intracellular water is not withdrawn. On the other hand, NaCl and sucrose remain outside such cells, withdraw water and evoke neuronal responses, which lead to thirst and ADH release. An action of the solutes on salivation, via the ventricular system, was also ruled out. There was no differential effect between intracarotid NaCl, sucrose and urea on the one hand, which have been shown to withdraw water from the ventricles (and concentrate its Na+) and glucose on the other hand, which did not withdraw water (McKinley et al., 1978). Continuous infusions of NaCl into the circulations of the brain and the parotid glands indicated that the inhibition of parotid secretion mediated centrally was achieved well within physiological limits of hypertonicity (threshold was 5 mOsm kg- 1) whereas that mediated by the gland was not (threshold was > 42.4 mOsm kg- 1). Since

46

W.L. Grovum and J.S. Gonzalez

the central osmosensitive neurons could be expected to be functional in day-to-day life, it was important to locate them for further study. The central osmosensitive site inhibiting parotid secretion was sought by injecting hypertonic saline into the circulation of the left half of the sheep brain (via the left mandibular artery), noting the inhibition, and then removing 5 mm slices of left brain from the front end until the inhibition was lost (Grovum, 1998). The inhibition was reduced largely when the ninth slice was removed which exposed the cranial 2 mm of the pituitary gland. It was assumed that the effect was due to the removal of many of the osmosensitive neurons. To pinpoint them, the brains in other sheep were stimulated with a concentric electrode in the region of the ninth slice. The rationale was that the osmosensitive neurons should inhibit parotid secretion whether they were excited by hypertonic solutes or electrical stimulation. Figure 3.3 illustrates such a response in the right gland to electrical stimulation on the right. On average, stimulating sites from 5 mm above the base to the base of the right brain, 13.7 mm lateral to midline and 1.9 mm rostral to the pituitary gland (29.5 mm rostral to earbar zero) inhibited right parotid secretion. Left parotid secretion was affected variably due to a variety of connections between the right and left cerebrum. This putative osmosensitive site (in the ventro-medial amygdala) inhibiting parotid secretion was found bilaterally and is new. It may help to protect body fluid levels along with the osmoreceptors in the OVLT involved in thirst and ADH release. Efferent activity in the parotid nerve was observed to decrease in multi-unit preparations after hypertonic NaCl injections into the carotid artery, but not after identical injections into the jugular vein. It appeared that the increase in plasma osmotic pressure from the carotid injections was detected in the brain and that, as a result, the parotid ‘centre’ in the medulla was inhibited. The decreased parasympathetic activity to the parotid gland then reduced or inhibited parotid secretion.

Fig. 3.3. Inhibition of right parotid secretion (right – ||| represents drops) by electrical stimulation of the right ventro-medial amygdala (from 5 mm above the base of the brain at earbars +26.7 mm, 13 mm to the right of midline) in sheep 272. Parotid secretion on the left and blood pressure are also shown. The top trace shows when stimuli (15 V, 40 Hz, 1 ms, 8 s every 30 s) were delivered through the concentric electrode and the numbers indicate the distance of the electrode tip from the base of the brain.

Control of Salivation and Motility of the Reticulorumen

47

The specialized osmosensitive system described above, which apparently inhibited saliva secretion to help maintain fluid balance, has manifested itself in ruminants, rats and humans. It is particularly important for ruminants because they secrete voluminous quantities of saliva compared with other species, probably to facilitate the processing of roughage particulate matter in the reticulorumen and the emptying of undigested residues into the lower gut. Dehydration reduces their saliva production (Silanikove and Tadmor, 1989) and their food intakes (Silanikove and Tadmor, 1989; Langhans et al., 1995) possibly because the associated systemic hypertonicity inhibits parotid as well as non-parotid salivation (Warner and Stacy, 1977). High-producing dairy cattle in hot arid environments have a threefold problem with water balance because they lose large quantities of water in the circulation to milk and saliva (temporary but none the less a loss) and, in addition, they must use water to keep cool. The problems are compounded when elevated body temperatures also reduce motility of the reticulorumen, slow emptying and increase rumen fill (Silanikove, 1992). Making the blood of rats hypertonic essentially abolished saliva secretion and diminished their heat tolerance because they cool themselves by spreading saliva on their body surfaces (Stricker and Hainsworth, 1970). A number of small fur-bearing animals do likewise. A similar mechanism may also exist in people as injections of 5% NaCl intravenously evoked thirst and reductions in salivary flow sensed as ‘dry mouth’ (Holmes and Gregersen, 1947). The phenomenon of postprandial hypertonicity is well established in ruminants. For example, osmotic pressures of plasma increased by 15–22 mOsm kg- 1 after meals in the work of Stacy and Warner (1966). This is likely to be exacerbated by feeding more concentrated diets. The dilemma is that, when high-producing ruminants need the buffering capacity of saliva the most to deal with an acidotic rumen (after meals), saliva secretion may in fact be inhibited by hypertonicity in blood. Feeding buffers to alleviate acidosis is common but questionable because the cations from the buffers, when absorbed, may elevate plasma osmotic pressure further, thus exacerbating the postprandial hypertonicity and causing further reductions in salivation. In addition, the ions from the buffer will elevate osmotic pressures in the reticulorumen fluid and hence, may reduce food intake (Grovum and Bignell, 1989). An alternative solution to acidosis in ruminants is to discover how to increase their intrinsic buffer supply by stimulating parotid secretion.

Investigations into brainstem control of gastric motility Review of background Involvement of the brain The role of the brain in motility of the reticulorumen was evident when stimulation of the peripheral cut end of the vagus nerve made it contract (Marschall, 1910 as cited by Andersson et al., 1959). When cyclic motility was eliminated by vagotomy (Mangold and Klein, 1927 as cited by Clark, 1953) but preserved in decerebrate sheep (Iggo, 1951), it was clear that the brainstem caudal to the intercollicular plane was important in programming motility sequences. The remaining brainstem was said to contain a ‘reticuloruminal motor centre’ which could be activated by distending the reticulum or

48

W.L. Grovum and J.S. Gonzalez

by stimulating the central cut end of one vagus (Iggo, 1951). Dussardier and AlbeFessard (1954) were even able to maintain cyclic motility of the reticulorumen with reticular distension and to stimulate it reflexly by vagal stimulation when the decerebration was made 5 mm below the junction of the pons and the medulla. Brain stimulation Contractions of the rumen were evoked by stimulating the interior of the brainstem electrically from midline to the lateral edge and from 2 mm caudal to 6 mm rostral to obex (Bell and Lawn, 1955). This site traversed many nuclei and fibre tracts and involved much of the reticular formation. Later, Andersson et al. (1958) elicited contractions of the reticulorumen in conscious goats by stimulating next to midline considerably below the dorsal motor nucleus of the vagus from obex - 1 mm to obex +3 mm. They argued that this nucleus might contain the ‘pace-maker’ for reticuloruminal motility even though stimulations close to it inhibited contractions. Curiously too, most of the points stimulated by Bell and Lawn (1955) which evoked contractions were outside this nucleus, and the two points within it were associated with oesophageal contractions, not with contractions of the rumen. Dussardier (1960) made the reticulum, but not the rumen, contract when he stimulated below the surface of the medulla in medial planes from obex - 2 mm to obex +1mm (the data can be seen in his Table No. 5). He postulated that circuitry existed in the reticular formation which initiated contractions of the reticulum. The stimulation of efferent fibres did not appear to account for many of these contractions as they were evoked from sites well below the fibre tracts from the dorsal vagal motor nucleus to the vagus (see Figure 61 in Dussardier, 1960). Dussardier (1960) further differentiated evoked contractions from the ability of brainstem stimulations to excite the rate circuit and affect the frequency of cyclic contractions. The loci which increased and decreased rate were distributed from 5 mm caudal to 5 mm rostral to obex over about two-thirds of the brainstem mass. The only clear groupings that may have constituted sites for decreasing rate were located first near the surface surrounding the dorsal vagal motor nucleus 3–4 mm caudal to obex and secondly, perhaps dorsolateral to the nucleus between 2 and 3 mm rostral to obex. This agreed with Andersson et al. (1959) who had arrested cyclic motility in conscious goats by stimulating the brainstem dorsolateral to this nucleus from 1 mm caudal to 2 mm rostral to obex. Furthermore, Howard (1970) found an inhibitory area in the interior of the medulla 2 mm from midline extending from the midpoint of the dorsal vagal motor nucleus cranially and ventrally. The abundance of inhibitory effects elicited by stimulation in and around the nucleus may explain why stimulating it directly failed to evoke contractions of the reticulorumen in the work of Howard (1970). The inhibition in some cases was short-lived as it was followed immediately by a reticular contraction (Howard, 1970). The dorsal vagal motor nucleus appears to be the origin of final efferent pathways to the different structures in the reticulorumen. The gastric motor fibres appear to originate in this nucleus (Dussardier, 1960: Figure 61; Howard, 1970) and bilateral lesions in these nuclei rostral to obex have abolished all motility in the reticulorumen (Beghelli et al., 1964 as cited by Howard, 1970). Single-unit recording Beghelli et al. (1963) used rather large electrodes (30 µm at the tip) to record bursts of electrical activity from the dorsal nucleus of the vagus before and during reticular con-

Control of Salivation and Motility of the Reticulorumen

49

tractions in young lambs (12–18 kg). However, they concluded that the centres initiating contractions may be located in the reticular formation as postulated by Dussardier (1960). This was because there was a delay of up to 10 s between the start of faradic stimulation of the central cut end of the vagus and records of bursts of electrical activity from the vagal nucleus, which preceded the evoked reticular contractions. Leek and Harding (1975) localized the ‘gastric centres’ in the dorsal vagal nucleus and up to 1 mm dorsal and lateral to it in the medulla from 2 mm caudal to 6 mm rostral to obex and 1.5–2.5 mm lateral from midline. The gastric motor neurons were found 2–2.5 mm below the surface at points 1 and 2 mm rostral to obex. However, this site was tiny compared with the sites where stimulations evoked contractions of the reticulorumen.

Surface sites from which reticular and ruminal contractions were evoked separately Stimulating the surface of the brainstem evoked contractions of the reticulum at one site (from 2 mm caudal to 2 mm rostral to obex and 1.8–5.6 mm lateral to midline) and the rumen at another (6.0–8.0 mm rostral to obex and 1.5–3.5 mm lateral to midline). These effects are illustrated in Fig. 3.4 and were first reported by Grovum and Gonzalez (1998c). Howard (1970) applied unipolar stimulation superficially to the

Fig. 3.4. Examples of stimulation on the surface of the brainstem, which evoked contractions of the reticulum only or of the rumen only, in sheep. The coordinates of the electrode tip are given relative to obex at 0 in the rostral–caudal or anterior–posterior stereotaxic plane (A–P – positive numbers go in the rostral direction) and 0 in the lateral plane. A concentric electrode was used (8 V, 1 ms, 40 Hz, 4 s). These results show a ‘reticular’ site lying 0–3 mm lateral and - 2 to +1 mm rostral to obex and a separate ‘ruminal’ site at 1–3 mm lateral and +6 to +8 mm rostral to obex.

50

W.L. Grovum and J.S. Gonzalez

medulla but did not find such responses. The tentative interpretation is that the gastric centres may receive inputs from these surface satellite sites to determine the amplitude and form of one of three reticular and of one of two ruminal contraction types seen during mixing, rumination or eructation.

Sites in the interior of the brainstem from which only ruminal contractions were evoked The data in Fig. 3.5 illustrate sites at considerable distances below the surface of the brainstem where stimulation evoked only ruminal contractions in 11 sheep. Most sites were between 4 and 8 mm rostral to obex (Grovum and Gonzalez, 1998c). There were once again large differences between sheep. The data are evidence for a second satellite site integrated with the ‘gastric centres’ to determine ruminal contractions. However, the extent of this ruminal site needs clarification. Stimulations nearby in the interior of the medulla usually evoked ruminal and reticular contractions together. The stimulations evoking the responses in Fig. 3.5 may therefore have identified only the edges of those ruminal sites which protruded beyond the overlapping reticular sites. The true ruminal sites may then be much larger than shown in Fig. 3.5. According to this

Fig. 3.5. A lateral view of the sites in the brainstem of individual sheep where stimulation with a concentric electrode evoked contractions of the rumen but not of the reticulum. The coordinates of the electrode tip are given relative to obex at 0 in the rostral–caudal or anterior–posterior stereotaxic plane (A–P; positive numbers are rostral to obex) and to the surface of the brain at 0 in the vertical plane. All points in the lateral plane in each sheep have been collapsed into the sites depicted.

Control of Salivation and Motility of the Reticulorumen

51

explanation, the neuronal circuits for the reticulum and the rumen were largely intermixed but functionally separate as seemed to be the case for the neighbouring salivary centres.

Two interior brainstem sites from which only reticular contractions were evoked One of these sites was usually oriented laterally close to obex (0.3–5.1 mm rostral to obex and 2.7–7.5 mm lateral to midline) and was 1.3–8.0 mm below the surface (Grovum and Gonzalez, 1998c). Stimulations alternating between this site and adjacent regions evoked either just reticular or reticular and ruminal contractions together. This is illustrated in one sheep in Fig. 3.6 even though the site was atypical in that it was oriented more in the cranial caudal direction than laterally as in most other sheep. The points marked with an asterisk indicate loci where just the reticulum contracted. Further, if one looks at the loci with rostral–caudal or anterior–posterior (A–P) stereotaxic readings between - 1 and +4, where the reticulum and the rumen contracted together, it is clear that the reticular contractions were often at least five times stronger than the ruminal contractions (e.g. 7 mm right, obex - 1 and depths 5–7 mm). Whereas, at obex +5 and +6, the difference was much smaller or even reversed (in second last

Fig. 3.6. Stimulation of the brainstem in sheep in the interior of the medulla (M) near obex evoked contractions of the reticulum (Re) alone (*). The rumen (Ru) may have contracted by itself also, albeit weakly (see 4th and 3rd last columns). A concentric electrode was used (8 V, 1 ms, 40 Hz, 4 s). The coordinates of the electrode tip are given relative to obex at 0 mm in the rostral–caudal or anterior–posterior (A–P) and lateral (L) stereotaxic planes and to the surface of the brain at 0 mm in depth (D). The top line indicates when stimuli were applied. The last column on the right shows the effects of stimulating the vagal rootlets inside the cranium with the electrode tip.

52

W.L. Grovum and J.S. Gonzalez

column on the right, the reticular strength was approximately one-third that of the ruminal strength). This may indicate a transition from a reticular site near obex to a ruminal site between 4 and 8 mm rostral to obex in accordance with Fig. 3.5. A careful inspection of the third and fourth last columns of Fig. 3.6 shows that some stimulations between 5 and 7 mm rostral to obex evoked weak ruminal, but no reticular contractions. The independence of the reticular and ruminal contractions is also evident in the middle of Fig. 3.6 at 5 mm lateral, A–P +6 and 4–6 mm depth. At a depth of 4 mm, the reticular and ruminal contractions were of equal size whereas at 5 and 6 mm depths, only reticular contractions were evoked. The contractions at 4 mm depth may therefore have resulted from stimulating functionally separate but intermixed circuits for the reticulum and the rumen. The last column on the right shows the effect of stimulating a point on the vagal rootlets within the cranium. Reticular contractions were also elicited on their own in five sheep at another location cranially distant from obex (8.5–15.4 mm; Grovum and Gonzalez, 1998c), beyond the most cranial sites from which contractions were evoked by Bell and Lawn (1955) (6 mm); and Dussardier (1960) (5 mm). This site was 4.5–9.3 mm lateral from midline and was 2.0–8.0 mm below the surface. Figure 3.7 illustrates representative data from one sheep in which most of the reticular contractions were evoked 8–14 mm rostral to obex. At the end of the experiment on this sheep both rumen and reticular contractions of normal strength were produced by stimulating the medulla just rostral to obex. The rumen was therefore capable of contracting but it did not respond at the loci shown in Fig. 3.7.

Sites controlling the lips of the reticular/oesophageal groove and the reticulo-omasal orifice A long air-filled balloon sutured loosely between the lips of the groove and a small pearshaped balloon dilated in the reticulo-omasal orifice of decerebrate sheep were monitored during brainstem stimulation to identify sites both on the surface and in the interior of the medulla which made these structures contract (Grovum, 1999). In one sheep out of four, three different combinations of groove and orifice responses existed to stimulation of the surface of the brainstem. First, both the groove and the orifice contracted in an oval area oriented caudo-laterally from 1 mm right (R) of midline at obex +6 mm at one end of the oval to 5R, obex +3 at the other. Laterally, in an adjacent oval area (response 2; from 3R, obex +7 to 9R, obex +3), the groove contracted and the orifice relaxed as would be expected during suckling. Finally, stimulations evoked only orifice dilations (response 3; balloon pressure decreased) in a large irregular area centred around a spot 8 mm lateral to obex. From the literature, sites would be expected which make the lips of the groove contract and the orifice either constrict or dilate because efferent vagal stimulation produced such responses (Titchen and Newhook, 1975). It may be that one groove site spanning responses 1 and 2 gave rise to the groove contractions and that one other site spanning responses 2 and 3 governed the orifice dilations. Stimulation where the sites overlapped caused both effects (response 2). However, the normal interactions between these sites are unknown. A third surface site in this sheep may have involved orifice constriction.

Control of Salivation and Motility of the Reticulorumen

53

Fig. 3.7. Stimulating the interior of the brainstem 8–14 mm rostral to obex evoked contractions of the reticulum but not of the rumen (representative data from one of five sheep). Obex was at 0 mm in the rostral–caudal or anterior–posterior (A–P) stereotaxic plane and at 0 mm lateral (midline). The vertical plane was illustrated with 0 mm being on the base of the brain because the surface was arbitrary due to the cerebellum having been removed by suction. Each of the five sagittal sections shows the magnitude of the reticular contractions as vertical bars with length being proportional to strength. Where the rumen contracted (7 mm right), the strength is indicated by a horizontal bar.

Brainstem site from which rumination and hypermotility has been evoked Rumination was evoked by stimulating the medulla in conscious sheep a few mm below the surface just dorsolateral to the dorsal nucleus of the vagus (from obex +2 mm to obex - 1 mm in the rostral–caudal plane; Andersson et al., 1959). This may have arisen from stimulation of epithelial receptor input pathways to the brainstem. Motility of the reticulorumen was inhibited at the same time. A high frequency of cyclic motility (5 cycles min⫺1) followed this inhibition. A similar result, except for the rumination, was obtained by Howard (1970) in halothane-anaesthetized sheep. Whether the marked stimulation of frequency had anything to do with the rate circuit of Leek and Harding (1975) is not known. Dussardier (1960) found that excitatory loci for cyclic motility were scattered throughout the brainstem ventro-laterally from the vagal nucleus. Electrical stimulation of the medulla near obex (1.8–4.1 mm lateral and from 0.6 mm caudal to 2.0 mm cranial) in anaesthetized sheep caused gas to be eliminated

54

W.L. Grovum and J.S. Gonzalez

from the reticulorumen (Grovum and Gonzalez, 1998a). This site was thought to be an eructation centre and was linked in later work (W.L. Grovum, unpublished data) to relaxation of the upper oesophageal sphincter. Nevertheless, the loss of gas could also have been simply associated with relaxation of this sphincter because the stimulation activated circuits involved with rumination or emesis. An emetic area was located by Andersson et al. (1959) near the surface of the medulla at obex +1 and obex +2 mm, dorsal to their ‘rumination’ area. Clearly, more work is required to understand the function of the site reported by Grovum and Gonzalez (1998a). A forebrain involvement in rumination is certain since it was evoked in conscious sheep by stimulating the ventrolateral part of the anterior hypothalamus electrically (Andersson, 1951; Larsson, 1954). Rumination could be stimulated with pentagastrin injections into the ventricular fluid (Honde and Bueno, 1984); it occurred as a conditioned response to milking stimuli in the goat (Andersson et al., 1958); and lobotomy and lesions in the ventral forebrain rostral to the optic chiasma increased activity to 24 h day- 1 in one sheep (Clark, 1953). The latter may be the result of removing a brake on the rumination centre, wherever it is, or of altering afferent input to that centre by altering descending inhibition (Urabe et al., 1968). The present authors are not aware of rumination being reported in decerebrate preparations but eructation can be induced (Titchen and Reid, 1965).

A putative mechanism for brainstem control of reticuloruminal motility The framework in Fig. 3.8 is highly speculative but it accommodates a diverse body of experimental findings. The main concept advanced is that separate satellite sites in the brainstem may initiate the separate contractions of the reticulum and the rumen during mixing, eructation and rumination by channelling their outputs in a coordinated sequence through the gastric centres. The following situations are addressed: 1. To initiate a normal motility cycle, A-type interneurons of the form and amplitude circuit in the gastric centres excite their respective motor neurons for the reticulum and the rumen at intervals of about 1 min (Harding and Leek, 1971; Leek and Harding, 1975). This appears to occur through disinhibition when the C-type interneurons (inhibitory and tonically active) are in turn inhibited by the B-type interneurons from the rate circuit. The A interneurons and the motor neurons are then released to discharge impulses in proportion to the balance of excitatory and inhibitory inputs being received from the rest of the brain and from the periphery. Further, it is not known if the inhibitory effects of stimulation on motility reported by Andersson et al. (1959), Dussardier (1960) and Howard (1970) are related to the inhibitory C interneurons of Harding and Leek (1971). 2. The three reticular contractions during rumination may originate from the three different satellite sites in the brainstem (one on the surface and two in the interior apart from the gastric centres) where stimulation evoked contractions of the reticulum only (see left half of Fig. 3.8). The first contraction associated with rumination would have to involve the rumination centre. There is no basis presently for linking specific contractions with specific sites.

Control of Salivation and Motility of the Reticulorumen

55

Rumen contraction sites

Reticulum contraction sites 1st of 2 2nd of 2 (mixing) (mixing) –

S2



C

– Delay

(Mixing) –

S3

– C

B

Regurgitation contraction +

+

+

+

+

Eructation centre

M

Rumen amplitude circuit of Leek and Harding (1975)

A M – Reticulum amplitude Rate circuit of Leek and Harding (1975) circuit of Leek and Harding (1975) C

S2

A

– –

S1 Delay

Synaptic delay

Rumination centre + S1

(Eructation) –

Vagal fibre Wall of reticulum

Fig. 3.8. A putative system in the brainstem of ruminants for controlling contractions of the reticulorumen during mixing, rumination and eructation (see the text for the details). There are no experimental data linking the three ‘reticular’ sites to any of the reticular contractions during mixing or rumination. Similarly, there are no data linking the two ‘ruminal’ sites to ruminal contractions during mixing and eructation.

3. The two ruminal brainstem sites (on surface and in the interior) may act independently to trigger ruminal contractions during either mixing or eructation (see top right of Fig. 3.8) but once again, there are no data to link a specific contraction to a specific site. The contraction during eructation would have to be activated by an eructation centre located elsewhere in the brainstem. Separate sites have not been identified for the dorsal and ventral sacs of the rumen. Furthermore, there is no basis for the backward and forward moving contractions in the dorsal sac during mixing and eructation respectively. 4. The ‘early’ motor neurons of Harding and Leek (1971) may act like a final common pathway to the reticulum as their discharges occurred in two phases, one of low frequency associated with the weaker first contraction in mixing and another of greater frequency associated with the stronger second reticular contraction. Accordingly, the scheme in Fig. 3.8 has the outputs from all reticular satellite sites routed through one set of motor neurons for the reticulum. 5. The ‘late’ motor neurons of Harding and Leek (1971) and Leek and Harding (1975) for the rumen may also act as a final common pathway as they were active during both mixing and eructation. This is accommodated in Fig. 3.8 by having the outputs from two satellite sites for the rumen routed through the A interneurons and the motor neurons for the rumen. 6. The amplitude of the second reticular contraction during mixing was inhibited more than the first as a result of stimulations starting from the midpoint of the dorsal nucleus of the vagus (2 mm from midline in the region of obex) in a line going in a

56

W.L. Grovum and J.S. Gonzalez

cranial–ventral direction toward the nucleus ambiguous (Howard, 1970). This could result if the satellite site for the second reticular contraction was inhibited rather than the A type interneurons or the motor neurons in the gastric centres. The satellite site for the first reticular contraction could then continue to operate normally. 7. Cyclic motility, which is generated in anaesthetized sheep, often starts with a single rather than a bi-phasic reticular contraction. This could be accounted for if the anaesthetic selectively inhibited one of the two satellite sites on the top left in Fig. 3.8 associated with mixing in the reticulum. 8. The strength of secondary dorsal sac contractions and their associated Type A activity exceeded that of primary contractions considerably (Iggo and Leek, 1967; Harding and Leek, 1971). Secondary contractions are more resistant than primary to drugs like xylazine (Grovum, 1986). The possibility of separate satellite controls for ruminal contractions in mixing and eructation could account for such observations. 9. The fact that ruminal motility is more depressed by anaesthetics (Iggo and Leek, 1967; Harding and Leek, 1971) and various abdominal stimuli (Titchen, 1960) than reticular motility, could be explained by a differential susceptibility of satellite sites. 10. The neural circuits in the brain which give rise to the synaptic delays, shown in Fig. 3.8, and produce a sequence of contractions in the reticulorumen have not been studied. The delays, however they are organized, must be variable as the mixing cycle is completed much faster when the stomach is empty than when it is full. 11. Research is needed to link the rate circuit of Leek and Harding (1975) to stimulation sites that markedly increased the frequency of cyclic motility.

Conclusions regarding salivation and reticuloruminal motility To improve our understanding of brain–gut interactions in ruminants, recordings of single-unit neural activity in the salivary and satellite motility sites will be a necessary but daunting task. Ultimately, this should help to improve rumen function and hence deal with the digestive problems associated with intensive production systems. When recordings are made from areas of the brain involved in controlling food intake, new ways may be found to increase roughage intakes and hence the productivity of ruminants at pasture.

References Andersson, B. (1951) The effect and localisation of electrical stimulation of certain parts of the brain stem in sheep and goats. Acta Physiologica Scandinavica 23, 8–23. Andersson, B., Kitchell, R. and Persson, N. (1958) A study of rumination induced by milking in the goat. Acta Physiologica Scandinavica 44, 92–102. Andersson, B., Kitchell, R.L. and Persson, N. (1959) A study of central regulation of rumination and reticulo-ruminal motility. Acta Physiologica Scandinavica 46, 319–338. Beghelli, V., Borgatti, G. and Parmeggiani, P.L. (1963) On the role of the dorsal nucleus of the vagus in the reflex activity of the reticulum. Archives Italiennes de Biologie 101, 365–384. Bell, F.R. and Lawn, A.M. (1955) Localization of regions in the medulla oblongata of sheep associated with rumination. Journal of Physiology London 128, 577–592.

Control of Salivation and Motility of the Reticulorumen

57

Carr, D.H. (1984) The regulation of parotid and submandibular salivary secretion in sheep. Quarterly Journal of Experimental Physiology 69, 589–597. Clark, C.H. (1953) The nerve control of rumination and reticulo-ruminal motility. American Journal of Veterinary Research 14, 376–384. Denton, D.A. (1957) A gregarious factor in the natural conditioned salivary reflexes of sheep. Nature 179, 341–344. Dussardier, M. (1960) Recherches sur le contrôle bulbaire de la motricité gastrique chez les ruminants. Doctoral thesis, Faculty of Science, University of Paris, France. Dussardier, M. and Albe-Fessard, D. (1954) Quelques propriétés du centre vagal contrôlant l’activité réflexe de l’estomac des ruminants. Journal de Physiologie 46, 354–357. Fitzsimons, J.T. (1989) Bengt Andersson’s pioneering demonstration of the hypothalamic “drinking area” and the subsequent osmoreceptor/sodium receptor controversy. Acta Physiologica Scandinavica 583, 215–225. Grovum, W.L. (1986) The control of motility of the ruminoreticulum. In: Milligan, L.P., Grovum, W.L. and Dobson, A. (eds) Control of Digestion and Metabolism in Ruminants. Reston book, Prentice Hall, Englewood Cliffs, New Jersey, pp. 18–40. Grovum, W.L. (1998) A putative osmosensitive site in the forebrain inhibiting parotid secretion in sheep. Journal of Physiology London 509, 8P. Grovum, W.L. (1999) Stimulation of the brainstem evoked opening and closing of the reticuloomasal orifice and closure of the reticular/esophageal groove in sheep. South African Journal of Animal Science 29(ISRP). Grovum, W.L. and Bignell, W.W. (1989) Results refuting volatile fatty acids per se as signals of satiety in ruminants. Proceedings of the Nutrition Society 48, 3A. Grovum, W.L. and Gonzalez, J.S. (1994) Centres in the brain stem controlling the secretion of parotid and mandibular saliva in sheep. Proceedings Society Nutritional Physiology 3, 76. Grovum, W.L. and Gonzalez, J.S. (1998a) A putative eructation centre in the sheep brain. European Journal of Neuroscience 10 (Suppl. 10), 250 (Abstract 97.01). Grovum, W.L. and Gonzalez, J.S. (1998b) Stimulation of the brain’s frontal lobe elicited body movements and parotid secretion in sheep. Journal of Animal Science 76 (Suppl. 1), 357. Grovum, W.L. and Gonzalez, J.S. (1998c) The reticulum and the rumen contracted separately to brainstem stimulation in sheep. Journal of Physiology London, 511, 54P. Harding, R. and Leek, B.F. (1971) The locations and activities of medullary neurones associated with ruminant forestomach motility. Journal of Physiology London 219, 587–610. Holmes, J.H. and Gregersen, M.I. (1947) Relation of the salivary flow to the thirst produced in man by intravenous injection of hypertonic salt solution. American Journal of Physiology 151, 252–257. Honde, C. and Bueno, L. (1984) Evidence for central neuropeptidergic control of rumination in sheep. Peptides 5, 81–83. Howard, B.R. (1970) The dorsal nucleus of the vagus as a centre controlling gastric motility in sheep. Journal of Physiology London 206, 167–180. Iggo, A. (1951) Spontaneous and reflexly elicited contractions of the reticulum and rumen in decerebrate sheep. Journal of Physiology London 115, 74–75P. Iggo, A. and Leek, B.F. (1967) An electrophysiological study of single vagal efferent units associated with gastric movements in sheep. Journal of Physiology London 191, 177–204. Langhans, W., Rossi, R. and Scharrer, E. (1995) Relationships between feed and water intake in ruminants. In: Engelhardt, W. von, Leonhardt-Marek, S., Breves, G. and Giesecke, D. (eds) Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction. Ferdinand Enke Verlag, Stuttgart, pp. 199–216. Larsson, S. (1954) Hyperphagia from stimulation of the hypothalamus and medulla in sheep and goats. Acta Physiologica Scandinavica 32 (Suppl. 115), 1–40. Leek, B.F. and Harding, R.H. (1975) Sensory nervous receptors in the ruminant stomach and

58

W.L. Grovum and J.S. Gonzalez the reflex control of reticulo-ruminal motility. In: McDonald, I.W. and Warner, A.C.I. (eds) Digestion and Metabolism in the Ruminant. The University of New England Publishing Unit, Armidale, pp. 60–76. McKenzie, J.S. and Smith, M.H. (1973) Stereotaxic method and variability data for the brain of the merino sheep. Journal fur Hirnforschung 14, 355–366. McKinley, M.J., Denton, D.A. and Weisinger, R.S. (1978) Sensors for antidiuresis and thirstosmoreceptors or CSF sodium detectors. Brain Research 141, 89–103. Silanikove, N. (1992) Effects of water scarcity and hot environment on appetite and digestion in ruminants. Livestock Production Science 30, 175–194. Silanikove, N. and Tadmor, A. (1989) Rumen volume, saliva flow rate, and systemic fluid homeostasis in dehydrated cattle. American Journal of Physiology 256, R809-R815. Stacy, B.D. and Warner, A.C.I. (1966) Balances of water and sodium in the rumen during feeding: osmotic stimulation of sodium absorption in the sheep. Quarterly Journal of Experimental Physiology 51, 79–93. Stricker, E.M. and Hainsworth, F.R. (1970) Evaporative cooling in the rat: Effects of dehydration. Canadian Journal of Physiology and Pharmacology 48, 18–27. Titchen, D.A. (1960) The production of rumen and reticulum contractions in decerebrate preparations of sheep and goats. Journal of Physiology London 151, 139–153. Titchen, D.A. and Newhook, J.C. (1975) Physiological aspects of suckling and the passage of milk through the ruminant stomach. In: McDonald, I.W. and Warner, A.C.I. (eds) Digestion and Metabolism in the Ruminant. The University of New England Publishing Unit, Armidale, pp. 15–29. Titchen, D.A. and Reid, C.S.W. (1965) The reflex control of the motility of the ruminant stomach. In: Dougherty, R.W. (ed.) Physiology of Digestion in the Ruminant. Butterworth, Washington, pp. 68–77. Urabe, M., Tsubokawa, T. and Hamabe, N. (1968) Studies of the activities of vagal afferents in the medulla oblongata and thalamus with special reference to a central regulatory mechanism. Physiology and Behavior 3, 17–28. Warner, A.C.I. and Stacy, B.D. (1977) Influence of ruminal and plasma osmotic pressure on salivary secretion in sheep. Quarterly Journal of Experimental Physiology 62, 133–142.

II

Rumen Microbiology and Fermentation

4

Molecular Ecology and Diversity in Gut Microbial Ecosystems R.I. MACKIE1, R.I. AMINOV1, B.A. WHITE1 AND C.S. MCSWEENEY 2 1Department

of Animal Sciences and Division of Nutritional Sciences, University of Illinois at Urbana-Champaign, Urbana, Illinois, USA; 2CSIRO Tropical Agriculture, Long Pocket Laboratory, Indooroopilly, Queensland, Australia

Introduction The earliest convincing evidence for life derives from 3.8 billion-year-old rock formations which have yielded microfossils including cellular filamentous forms and sheathed colonies resembling blue–green bacteria (Schopf, 1993). Also, all basic evolution in terms of biochemistry and types of energy metabolism must have occurred earlier than 3.5 billion years ago during periods for which we have no direct evidence. Early evolution took place on earth in which most habitats (the atmosphere and oceans) were anaerobic. Thus, bacteria are nearly as old as the planet earth and events such as the oxygen forming atmosphere and even the age of dinosaurs are comparatively recent (Woese, 1994). Also, anaerobic habitats have existed continuously throughout the history of the earth, the gastrointestinal tract being a contemporary microniche (Fenchel and Finlay, 1995). The most obvious impact of fermentation in the modern world is on human and animal nutrition. The annual cellular production of prokaryotes based on population size (4.3 ¥ 1024 cells) and turnover (once daily) in the gut of domestic animals is 2 ¥ 1027, which is 2% of that in soil and only 0.1% of that in the oceans (Whitman et al., 1998). All animals, including humans, are adapted to life in a microbial world. The complexity of animal–microbe relationships varies tremendously, ranging from competition to cooperation (Hungate, 1976, 1984). The animal alimentary tract has evolved as an adaptation enabling the animal to secure food and limit consumption by other animals. This allows the retention and digestion of ingested food, followed by absorption and metabolism of digestion products, whilst feeding and other activities continue. Since microorganisms grow rapidly under favourable conditions in the gut they could become serious competitors for the animal’s food. This microbial challenge has modified the course of evolution in animals, resulting in selection for varied animal–microbe relationships. The evolutionary strategy in the first case has been to compete with the resident microbes and in the second to cooperate with them. The third case incorporates a combination of the first two avoiding some of the disadvantages of the © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

61

R.I. Mackie et al.

62

cooperation while still obtaining some benefit from fermentative digestion. These models of existing animal–microbe relationship are useful when trying to determine the evolution of mutualistic fermentative digestion in the gastrointestinal tract.

The first mammalian herbivores During the late Cretaceous and early Paleocene periods plant-eating mammals were frugivores, presumably because fruit can be more easily processed than foliage. Mammals did not invade the herbivorous niche until the middle Paleocene (Collinson and Hooker, 1991). Evolution of large size was a prerequisite for the exploitation of leaves because of the need for a longer residence time in the gut for bacterial fermentation to obtain sufficient nutrients from foliage and herbage. In the late Cretaceous, dinosaurs occupied the herbivorous niche although grazers were still absent (Farlow, 1987). The appearance of grazers in the Miocene is coincident with a similar radiation of grassland-forming grasses (Thomasson and Voorhies, 1990). Thus herbivore browsers first appear in the middle Paleocene but they did not become significant until the late Eocene (Collinson and Hooker, 1991). Frugivory declined first with the appearance of herbivore browsing followed by an increase in grazers in post-Miocene at the expense of herbivore browsers. The earliest herbivores were ground dwelling (LGMs, large ground mammals) and achieved their dietary specializations largely through evolution from already large, ground-dwelling frugivores or, in the Paleocene, by a size increase from small insectivorous ancestors (Collinson and Hooker, 1991). Large size limited them to the ground. Most browsing herbivores in other locomotor niches (SGMs, small ground mammals) changed their diet from frugivory without changing their locomotor adaptation. A period of nearly 30 million years existed in the vertebrate exploitation of leaves after dinosaur extinction and before the first few mammalian herbivores in the middle Paleocene. This was followed by expansion of herbivores in the late Eocene when climates cooled and more open vegetation became established. Hume and Warner (1980) published an excellent discussion on the evolution of microbial digestion in mammals. Since the fossil record provides no information on the morphology, physiology, biochemistry or microbiology of the gut, much of the knowledge must be deduced from what is known about present-day animals coupled with the fossil record of animals and their probable feedstuffs and is therefore highly speculative. Microbial digestion surely arose long before mammals evolved. Large luminal populations of microbes develop in regions of the gut with relative stasis where retention time of digesta allows adequate microbial growth. In most deliberations only nutritional contributions to the host animal are considered. These are based on digestion of the plant cell wall by cellulases and hemicellulases provided by the microbial partner, the synthesis of microbial protein from poor-quality dietary proteins and non-protein nitrogen mainly via ammonia as precursor, and the synthesis of B-vitamins and vitamin K. However, little consideration has been given to the protection provided by foregut fermentation resulting in transformation or modification of phytotoxins and mycotoxins in the diet (Guthrie, 1984; Mackie, 1987; McSweeney and Mackie, 1997). Other contributions not normally considered in these deliberations are immunological, physiological and protective (Gaskins, 1997; Rolfe, 1997; Tannock, 1997). These arguments

Molecular Ecology and Diversity in Gut Microbial Ecosytems

63

also support the theory that the development of foregut fermentation must have come after an initial development of the hindgut and that all foregut fermenters should have some fermentation in the hindgut (Hume and Warner, 1980; Langer, 1991). The above mentioned adaptations, modifications and specializations in anatomy and physiology of animals for herbivory would be ineffective in the absence of endosymbiotic microbes to digest the plant cell wall. This raises the question of the actual mode of acquisition of these endosymbionts by their hosts. Since it is unlikely that the fossil record will ever provide direct clues for resolution of this issue, Hotton et al. (1996) have hypothesized that microbes were picked up by detritivorous animals foraging in plant litter and those that could survive in the gut environment have assumed a role in the digestive processes of the host. However, it is equally plausible that ingested insects, especially herbivores that harbour such bacteria in their own guts, provided the original source for fermentative endosymbionts (Sues and Reisz, 1998). This concept is compatible with contemporary studies on acquisition of gut microbiota by neonates which document the importance of diet and environment in the early development of microbial populations (Mackie et al., 1999). It is also consistent with development of hindgut fermentation and omnivory that included consumption of plant material prior to development of foregut fermentation. This consumption of insect herbivores by mammalian hosts may be involved in coevolution, the continual process of evolutionary change in the synthesis of secondary compounds by plants followed by comparable detoxification mechanisms in animals that consume them (McSweeney and Mackie, 1997). The opportunity to examine the intestinal contents of an extinct mammal is rare. Recent excavations at two separate sites in the Great Lakes region of North America revealed assemblages of plant material preserved in late-Pleistocene pond sediments associated with skeletal remains of American mastodons (Mammut americanum).The plant material from these assemblages varied in degree of comminution but differed in texture and colour from surrounding sediment and resembled gut content from modern herbivores. The shape and size of the mass of plant material was consistent with the intestinal dimensions of extant elephants (Lepper et al., 1991). The plant material was dated to 11,500 years before present and the masses are thought to be remnants of the small and large intestines. General and selective media were used to cultivate and identify bacteria from the intestinal contents, bone-associated sediments and sediments located some distance from the remains (Rhodes et al., 1998). In all, 295 isolates were cultivated and 38 individual taxa identified. Sequencing of 16S rDNA was used to confirm the taxonomic positions of selected enteric and obligately anaerobic bacteria. Members of the family Enterobacteriaceae represented 41% of all isolates (Enterobacter cloacae was the most commonly identified isolate) in the intestinal masses. However, no Bacteroides spp. or expected intestinal anaerobes were recovered. In fact, the only obligate anaerobes recovered were clostridia and these were not from the intestinal masses. It was concluded that microbiological evidence supported other macrobotanical data indicating the intestinal origin of these plant masses but it was not possible to establish whether these organisms are direct descendents of the original intestinal microbiota (Rhodes et al., 1998). Although interesting, the results of this study are ambiguous and a more direct method of studying ancient bacteria is required. It is apparent, using rigorous DNA technology protocols, that a picture of the composition of the original

R.I. Mackie et al.

64

microbiota can be obtained at least under special circumstances which allow differentiation of ancient and modern DNA and that mammoths and similarly well-preserved bodies are attractive candidates for future palaeomicrobiological investigations (Rollo and Marota, 1999).

Molecular ecology and diversity The microbial community inhabiting the gastrointestinal tract is characterized by its high population density, wide diversity and complexity of interactions. All major groups of microbes are represented in the gut and it contains representatives of the three domains, Bacteria, Archaea and Eucarya, articulated by Woese et al. (1990). The rumen, the most extensively studied gut ecosystem, contains large numbers of bacteria (up to 1011 viable cells ml- 1 comprising 200 species), ciliate protozoa (104–106 ml- 1 spread over 25 genera), anaerobic rumen fungi (zoospore population densities of 103–105 ml- 1 divided into five genera) and bacteriophages (107–109 particles ml- 1) (Klieve and Swain, 1993; Hespell et al., 1997; Orpin and Joblin, 1997; Stewart et al., 1997; Williams and Coleman, 1997). Despite this vast amount of knowledge, the basic prerequisites for ecological studies, namely enumeration and identification of community members, have tremendous limitations. The two major problems faced by microbial ecologists studying the gastrointestinal ecosystem are the inevitable bias introduced by culture-based enumeration and characterization techniques and the lack of a phylogenetically-based classification scheme (Ward, 1989; Amann et al., 1990, 1994; Stahl and Amann, 1991; Stahl,1997). Modern molecular ecology techniques based on sequence comparisons of nucleic acids (DNA or RNA) can be used to provide molecular characterization while at the same time providing a classification scheme which predicts natural evolutionary relationships. In principle, nucleic acid probes can be designed to hybridize with a complementary target sequence and thus provide a complete description independent of the growth conditions and media used (Ward et al., 1992; Amann, 1995; Amann et al., 1995; Raskin et al., 1997). An example of the power of these modern molecular ecology techniques is provided by the analysis of 16S rRNA sequences (average length 1500 nucleotides). The highly conserved regions of the rRNA molecule can serve as primer binding sites for in vitro amplification by PCR (Ludwig et al., 1994). The more conserved regions are also useful, serving as targets for universal probes that react with all living organisms or for discriminating between broad phylogenetic groups such as the domains Archaea, Bacteria and Eucarya. The more variable sequence regions are more appropriate for genus, species and sometimes even strain-specific hybridization probes (Stahl and Amann, 1991; Odenyo et al., 1994a; Lin and Stahl, 1995). The use of molecular ecology techniques based on nucleic acid probes is likely to revolutionize our approach to microbial ecology in the gastrointestinal tract and will provide, not simply a refinement or increased understanding but, a complete description of the gastrointestinal ecosystem for the first time (Mackie and Doré, 1997; Raskin et al., 1997). In molecular ecology it is important to distinguish between identification, quantification and monitoring of activity or function. This information is highly dependent on the type of target nucleic acid employed and the conceptual and technical basis used

Molecular Ecology and Diversity in Gut Microbial Ecosytems

65

for nucleic acid probe design. Generally these can be divided into DNA-based methods employing empirically characterized probes and rRNA-based methods based on comparative sequence analysis for design and interpretation of ‘rational’ probes (Pace et al., 1986; Stahl and Amann, 1991; Ward et al., 1992; Stahl, 1993b; Raskin et al., 1997).

DNA-based studies of gut ecology The earliest nucleic-acid-based analyses of ruminal ecology were published more than a decade ago in order to study survival of bacteria reintroduced into the rumen. Attwood et al. (1988) reported on the use of a 1.9 kb randomly cloned, labelled fragment of genomic DNA from the laboratory strain Prevotella (Bacteroides) bryantii to determine survival time of this organism when introduced into the rumen. Results showed that strain B14 had a half-life of 9 h for in vitro mixed cultures but only 30 min following inoculation into the rumen. In fact, the organism dropped below the detection limit (2 ¥ 107 cells ml- 1 rumen fluid) within 3 h, most likely due to bacteriocin-like activity in fresh ruminal fluid. A number of studies have used randomly cloned DNA fragments targeting either plasmid or chromosomal DNA as hybridization probes. Streptococcus bovis, a normal inhabitant of the rumen, has been implicated in human colonic carcinoma. Whitehead and Cotta (1993) tested a cloned amylase gene as a DNA probe for rapid and accurate identification of S. bovis derived from human and bovine sources. In general, the DNA-based methods for studying gastrointestinal molecular ecology can have high specificity and sensitivity. However, it is necessary to identify those methods that require pure culture isolates (restriction fragment length polymorphism and ribotyping) and therefore have the disadvantage of cultural bias. In addition, DNA probes have also been used to investigate species diversity in the gastrointestinal tract including studies on Fibrobacter (Flint et al., 1990), Selenomonas ruminantium (Ning et al., 1991), and Butyrivibrio fibrisolvens (Mannarelli et al., 1990). Furthermore, these techniques are less suitable for detailed characterization of community structure of gastrointestinal ecosystems compared to rRNA techniques which employ rational probes designed within the framework of comparative sequence analysis and are therefore highly specific and span a greater taxonomic diversity.

RNA-based studies of gut ecology The principles and practice involved in 16S rRNA-based methods have been extensively reviewed (Pace et al., 1986; Sayler and Layton, 1990; Ward et al., 1992; Stahl, 1993a,b; Raskin et al., 1997). The use of small subunit rRNA methods, referred to as 16S rRNA-based methods, in gastrointestinal microecology has become well established over the past 10 years. These methods are also applicable to 23S rRNA which will probably be used more extensively in the future since it contains more information (3000 bp compared with 1500 bp for 16S rRNA). At present their use is limited because of the less extensive sequence collection available. Extensive comparative sequence analysis of 16S rRNA molecules representing a wide diversity of organisms shows that different regions of the molecule vary in sequence conservation.

66

R.I. Mackie et al.

Oligonucleotides complementing regions of universally conversed 16S rRNA sequence are used as universal probes while those complementing more variable regions of sequence are useful as selective probes targeting species, genus, or phylogenetic groups. The use of 16S rRNA-based methods for ecological studies in the gastrointestinal tract was first demonstrated by Stahl et al. (1988). Species-specific 16S rRNA-targeted oligonucleotide probes were developed to enumerate strains of Fibrobacter (Bacteroides) succinogenes and Lachnospira multiparus in the bovine rumen to monitor shifts in population abundance or changes in population activity in complex microbial communities. Subsequently, F. succinogenes was shown to be a genetically diverse yet phylogenetically coherent genus (Amann et al., 1992; Ogata et al., 1997). Culture-based enumeration of F. succinogenes was largely unsuccessful in the same study. These techniques formed the basis for further studies of bacterial populations in gut environments. Group-specific 16S rRNA-targeted oligonucleotide probes specific for bacterial, eukaryotic, and archaeal rRNA were used to quantitate these three groups in the gastrointestinal tract of various domestic animals (bovine, ovine, caprine and porcine). Bacterial, eukaryotic and archaeal populations varied from approximately 60 to 90%, 3 to 30% and 0.5 to 3%, respectively, in the gastrointestinal tract of most of the animals examined (Lin et al., 1997). By using different probes for classes of methanogenic bacteria, it was also shown that the predominant methanogens differed in different animal hosts (Lin et al., 1997). Moreover, these studies revealed a dominant archaeal group that was unaccounted for by the current methanogen-specific probes, suggesting another assemblage of rumen methanogens that have yet to be characterized (Lin et al., 1997). Krause and Russell (1996) developed probes for obligate amino-acid-fermenting ruminal bacteria in order to examine the effect of monesin addition on the levels of these bacteria and their role in amino acid deamination. Probes were developed for the other major ruminal cellulolytic bacteria Ruminococcus albus and Ruminococcus flavefaciens and used to study the dynamics of bacterial interactions during fermentation of cellobiose, cellulose and alkaline hydrogen peroxide treated wheat straw (Odenyo et al., 1994a,b). The results showed that 16S rRNA-targeted oligonucleotide hybridization probes were effective in quantifying specific bacteria in defined mixed cultures and provided useful information on bacterial competition during growth on insoluble substrates. Furthermore, this study revealed for the first time the production of a bacteriocin-like substance by a ruminal bacterium, a mechanism which may be used to compete for nutrients. These probes have also been used to study the competition for cellobiose or cellulose by R. albus, R. flavefaciens and F. succinogenes in chemostats under substrateexcess and substrate-limited conditions (Shi and Weimer, 1997; Shi et al., 1997). The ruminal bacterium Synergistes jonesii has attracted considerable interest due to its limited geographical distribution and its ability to degrade 3-hydroxy-4-1(H)-pyridone, the toxic principle of the tropical leguminous shrub Leucaena leucocephala (Allison et al., 1990, 1992). The 16S rRNA sequence of S. jonesii was not closely related to any among those bacteria so far characterized and was an ideal candidate to evaluate the use of oligonucleotide probes for tracking bacteria in the ruminal ecosystem. Radiolabelled and fluorescent-dye-conjugated probes were developed for quantitation of S. jonesii in a mixed-culture chemostat (McSweeney et al., 1993). Further ecological studies on transmission, colonization, persistence and population studies in vivo are still required. Surprisingly, a limited number of studies have used rRNA-based methods to study ruminal protozoal and fungal populations. Finlay et al. (1994) used a fluorescently

Molecular Ecology and Diversity in Gut Microbial Ecosytems

67

labelled oligonucleotide probe targeting the 16S rRNA of the Archaea to demonstrate that Entodinium species and Dasytricha ruminantium contained methanogenic endosymbionts outside digestive vacuoles. Exosymbiotic methanogens had been well documented previously, based on characteristic autofluorescence of these bacteria (Vogels et al., 1980; Stumm et al., 1982; Krumholz et al., 1983). Recent research using a small subunit (SSU) rRNA probe approach targeting ruminal methanogens revealed a taxon-specific association between protozoal and methanogen populations both in the rumen and a continuous culture fermentor system (Sharp et al., 1998). Methanobacteriaceae were the most abundant population in the rumen comprising 89.3% of total Archaea and 99.2% in the protozoal fraction. This value decreased to 54% of Archaeal signal after 48 h of fermentor operation and was correlated with the loss of protozoa from the system. In contrast, the Methanomicrobiales, the most abundant Archaeal population in this study, accounted for 12.1% of Archaeal signal in rumen fluid and was not detected in the protozoal fraction suggesting a free-living lifestyle. This group increased to 26.3% of Archaeal signal in fermentor content without protozoa. These studies suggest the importance, and perhaps specificity, of Methanobacteriaceae as symbionts of rumen protozoa. Initially, sequence analysis of the 18S rRNA genes from Polyplastron multivesiculatum (originally deposited as Entodinium simplex) and D. ruminantium was used to phylogenetically position these protozoa within the hydrogenosome containing protozoa (Embley et al., 1995). More recently, seven other rumen ciliate protozoa, as well as two additional isolates of P. multivesiculatum and D. ruminantium, have been analysed and these studies indicate that the rumen ciliates are monophyletic and fall into three distinct groups within the Class Litostomatea (Wright and Lynn, 1997a,b; Wright et al., 1997). Signature probes for Entodinium caudatum, Epidinium caudatum and P. multivesiculatum have been developed and are currently being analysed for use in the study of rumen microbial ecology (Wright et al., 1997). Also, comparative sequence analysis of the 18S rRNA genes was used to phylogenetically position Neocallimastix in the Chytridiomycete class of fungi (Doré and Stahl, 1991). We have recently determined the 18S rRNA sequences for four additional rumen fungi and these data support the previous comparative analysis and suggest that the anaerobic rumen fungi are monophyletic (Thill et al., 1997).

Recent advances and new approaches Sequence-based analysis of rumen microbial diversity Development of robust and simple DNA cloning techniques and the PCR have allowed higher resolution analyses of complex communities using SSU rRNA comparative sequence analysis (Hugenholtz and Pace, 1996). This approach has been strongly advocated by Pace (1997). This technique has not been widely applied to the gut ecosystem but studies of the termite gut (Okhuma and Kudo, 1996), the human colon (Wilson and Blitchington, 1996; Doré et al., 1998) and the rumen (Whitford et al., 1998; Tajima et al., 1999) have been published. Results of these studies demonstrate the high degree of genetic diversity in gut ecosystems. Whitford et al. (1998) prepared two libraries from strained rumen fluid obtained from dairy cows fed a total mixed

68

R.I. Mackie et al.

ration containing 26% lucerne, 30% maize silage and 35% concentrate. Sequences which clustered with Prevotella ruminicola represented the majority of clones (101 of 133 total sequences) isolated in each of the PCR sets. However, many members of this cluster represent phylogenetically distinct groups and are at least different species. Relatively few 16S rRNA sequences similar to the commonly isolated B. fibrisolvens were found in this study. Importantly, the majority of rDNA sequences analysed in this study represented novel bacterial diversity which has not yet been cultivated or isolated. A recent study (Tajima et al., 1999) described bacterial diversity by direct retrieval of 16S rDNA sequences in a culture-independent manner. Three SSU rDNA libraries were constructed; the first from total DNA extracted from strained rumen fluid of representative samples obtained before feeding from cows fed a mixed ration (lucerne–timothy hay and maize+barley based concentrate in a 4:1 ratio), the second from the remaining feed particles from the first pooled sample, and the third from strained rumen fluid of cows fed a high roughage (lucerne–timothy hay) diet. The three libraries containing almost full length SSU rDNA sequences (about 1.5 kb long) were completely sequenced and analysed (a total of 161 clones). Only 10 sequences (6.2%) could be identified, six as B. fibrisolvens, two as P. ruminicola, one as S. ruminantium and one as Succiniclasticum ruminis. For 34.6% of the sequences the similarity with database sequences ranged from 90 to 98% while for the remaining 59.2% the similarity was less than 90%. Because of the large bacterial diversity, few operational taxonomic units (OTUs) represented a large percentage of the clones. Phylogenetic placement of sequences from the mixed ration/rumen fluid library showed the following affiliations: low G+C Gram-positive bacteria (52.4%), Cytophaga– Flexibacter–Bacteroides (38.1%), Proteobacteria (4.7%) and Spirochaetes (2.4%). Values for the analysis of sequences from the solids remaining (library two) were low G+C Gram-positive bacteria (71.4%), Cytophaga–Flexibacter–Bacteroides (26.2%) and Spirochaetes (2.4%). Corresponding sequences from the third library (high roughage rumen fluid) were low G+C Gram-positive bacteria (44.2%), Cytophaga– Flexibacter–Bacteroides (50.6%) and Spirochaetes (3.9%). In addition, 418 randomly isolated colonies from a number of non-selective media were characterized and SSU rDNA sequences obtained. Results showed that 59.6% of strains were identified as P. ruminicola, 10.8% as Prevotella spp. and 0.8% as Bacteroides. The remaining 27.0 and 1.8% of isolates were affiliated with low G+C and high G+C Gram-positive bacteria, respectively. In contrast to the library-based analysis, the cultivation-based phylogenetic approach revealed close clustering with strains that have already been isolated, characterized and sequenced. Similar results for the human colonic ecosystem in terms of great diversity, large proportion of OTUs represented by single clones, and a large proportion of sequences distantly related to deposited sequences and so far uncultivated (Wilson and Blitchington, 1996; Doré et al., 1998). These approaches all provide useful and novel information but also have limitations which need to be recognized and resolved. The limitations relate to the extraction of nucleic acids from environmental samples, biases, artifacts associated with enzymatic amplification of nucleic acids and cloning of PCR products, and sensitivity and target site accessibility in whole-cell hybridization techniques. These biases and limitations have been well documented and reviewed (Wintzingerode et al., 1997; Head et al., 1998; Muyzer and Smalla, 1998). However, of importance is interpretation of information derived from molecular ecology studies (Stackebrandt, 1997). For example the 16S

Molecular Ecology and Diversity in Gut Microbial Ecosytems

69

rRNA sequence variations due to interspecific and intraspecific rRNA operon heterogeneity make analysis of clone libraries or gel electrophoresis patterns derived from environmental samples difficult to interpret (Hunter-Cevera, 1998). In addition, it is hard to draw conclusions about physiological and biochemical properties and the ecological role of unknown (and known) microbes. However, considerable progress in addressing and resolving these limitations is being made.

Application of denaturing gradient gel electrophoresis (DGGE) and temperature gradient gel electrophoresis (TGGE) in gut microbial ecology In order to study population structure and dynamics, genetic fingerprinting techniques that provide a pattern or profile of genetic diversity are needed. Recently DGGE of PCR-amplified rDNA fragments has attracted the attention of microbiologists and has been applied in a variety of environmental studies for the analysis of microbial communities (Muyzer et al., 1993). In DGGE, as well as in TGGE, DNA fragments of the same length but with different sequences can be separated. Separation is based on the decreased electrophoretic mobility of a partially melted double-stranded DNA molecule in polyacrylamide gels containing a linear gradient of denaturant (a mixture of urea and formamide) or a linear temperature gradient. Sequence variation within such domains causes the melting temperatures to differ and molecules with different sequences will stop migrating at different positions in the gel. DNA bands in DGGE and TGGE profiles can be visualized using ethidium bromide, SYBR Green I or silver staining which is more sensitive but also stains single-stranded DNA. This procedure allows direct identification of the presence and relative abundance of different species and provides a semi-quantitative estimation of the genetic diversity of microbial populations. DGGE and TGGE of PCR-amplified DNA fragments has been used widely in environmental microbiology to study community complexity, monitor population shifts, analyse enrichment cultures and isolation of bacteria, detect sequence heterogeneity of 16S rRNA genes, compare DNA extraction methods, screen clone libraries and determine PCR and cloning biases (reviewed by Muyzer et al., 1998; Muyzer and Smalla, 1998). This technique has had surprisingly limited application to the gut ecosystem most probably because of the complexity of banding patterns which makes it difficult to analyse. We have applied this procedure to pure cultures of ruminal fibrolytic bacteria F. succinogenes, R. albus and R. flavefaciens as well as other members of the genus Ruminococcus which may not be true members of this genus (R. bromii, R. gnavus, R. lactaris, R. obeum and R. torques). DGGE profiling of the nine R. flavefaciens strains showed that the PCR-amplified V3 region of the 16S rRNA gene from these strains all migrated to the same location in the DGGE gel. These results suggest a lack of sequence diversity in this region of the 16S rRNA gene from R. flavefaciens. In contrast, DGGE profiling of the nine R. albus strains suggested a high degree of sequence diversity in the PCR-amplified V3 region of the 16S rRNA gene from these strains. These results suggest that there is significant sequence diversity and phylogenetic heterogeneity within the V3 region of the 16S rRNA gene from R. albus, and that the V3 region of the 16S rRNA gene from R. albus SY3 has a much lower G+C content than other representatives of this species (Cann et al., 1996). All of the Ruminococcus DGGE profiles

70

R.I. Mackie et al.

differed from the profile for F. succinogenes, and mixtures of PCR-amplified V3 regions from the strains of F. succinogenes, R. albus, and R. flavefaciens were easily resolved in DGGE gel. Further studies with different dilutions of input DNAs from these strains showed that the DGGE technique is quantitative. These results show that DGGE can be used to differentiate between closely related bacterial strains. Thus this sensitive technique is highly suitable to the analysis of microbial diversity and population dynamics for the major fibrolytic bacteria from the rumen. We have also applied this procedure to rumen samples from steers fed different diets in a preliminary study to determine the utility of these techniques for the analysis of a complex microbial community (Kocherginskaya et al., 1997). Rumen samples were collected from four steers fed a medium-quality grass–legume hay at maintenance intake, and four steers fed a diet of 20% hay, 52% maize, 5% corn steep liquor, 3% minerals and 20% of maize byproducts. Rumen samples were harvested approximately one hour prior to feeding, passed through cheesecloth, and centrifuged. Total genomic DNA was isolated from cell pellets and used for amplification of either the V3 or V9 region of the 16S rDNA gene. When the different PCR profiles obtained from amplification of the V3 region of the 16S rDNA were compared for samples from those animals fed the medium-quality grass–legume hay diet, the patterns were remarkably similar. None the less, the DGGE profiles demonstrated at least 16 distinguishable bands, with five of them being more predominant than the others. Banding profiles obtained from rumen samples of animals fed the maize-based diet were different from those obtained from the medium-quality grass–legume hay diet. Profiles from each of the four maize-based-diet-fed animals also differed from each other. These results demonstrate the utility of this technique in describing the genetic diversity and population structure of the rumen community, both in vitro and in vivo. Bacterial diversity in human faeces was analysed using PCR-amplification of the V6–V8 regions of 16S rDNA by TGGE (Zoetendahl et al., 1998). Faecal samples from two individuals showed remarkably stable profiles over a period of at least 6 months and were unique for each individual. TGGE profiles derived from 16S rRNA (by reverse transcriptase PCR) and rDNA amplicons showed similar banding patterns although the intensities of bands with similar mobilities differed in some cases indicating a different contribution to the total active fraction of the predominant faecal bacteria. These results confirm that TGGE analysis of 16S rDNA amplicons, combined with cloning and sequencing of these amplicons, is a reliable approach to relative levels and abundance in complex microbial communities such as faecal bacteria. Bacterial genetic diversity in pig faecal samples was determined using DGGE (Simpson et al., 1999). Optimization of the protocol resulted in a doubling of product bands visualized in the gels. Unique and stable banding patterns were generated from faecal samples of pigs on different diets and of different ages as well as from lumenal and mucosal samples obtained from each gut segment between the stomach and colon. Analysis of the DGGE banding profiles using PHYLIP showed that the patterns grouped according to gut location and that lumenal and mucosal samples from each compartment had the highest similarity to each other (Simpson et al., 1999). DGGE can be applied effectively to monitor changes in bacterial populations and for evaluation of bacterial diversity. However, many of the specific gut populations are minor constituents and there are detection limits for populations comprising less than 1% of total template DNA and PCR amplification with semi-conserved primer sets (Muyzer and Smalla, 1998).

Molecular Ecology and Diversity in Gut Microbial Ecosytems

71

Whole-cell hybridization Whole-cell in situ hybridization with fluorescently-labelled oligonucleotide probes for studies in microbial ecology was first developed 10 years ago (DeLong et al., 1989; Amann et al., 1990). This technique was used successfully to analyse a wide variety of ecosystems (Amann et al., 1995, 1996). However, applications to analysis of the rumen ecosystem are limited. In short, the procedure involves cell fixation to permeabilize cells while maintaining their morphological integrity. Hybridization with a fluorescent probe to bind with complementary rRNA sequence is carried out on bacterial suspensions or after attachment to coated microscope slides. Following hybridization, the sample is washed to remove unbound probe and the sample viewed by epifluorescence microscopy (Amann et al., 1995). Recent developments that have improved our ability to address structure and function of microbial communities in situ have been reviewed (Amann and Kuhl, 1998). Localization of microorganisms on mucosal or cell surfaces is usually performed by classical histological and immunohistological techniques, which restrict identity of the bacteria involved to morphological features. Even when specific antibodies are available for in situ studies, the thick mucous layer above epithelial cells can block penetration of antibodies and extensive washing can remove the mucus layer displacing organisms. Poulsen et al. (1994) applied in situ 16S rRNA hybridization to investigate the microbiota of the large intestine of streptomycin-treated mice and were able to determine the spatial distribution of Escherichia coli in thin sections of intestinal tissue. This allows rapid detection of bacteria which may be difficult to cultivate and their relationship to other cells either host or bacterial. In situ growth rates determined by single-cell analysis of intracellular concentrations of DNA and RNA revealed that adherent and mucosal bacteria were growing with generation times of 30–80 min while those in the lumen were static (Poulsen et al., 1995). An advantage of the fluorescent in situ hybridization approach is the ability to both identify and enumerate single cells within a complex ecosystem with specific 16S rRNA-based oligonucleotide probes. Thus, six 16S rRNA-targeted oligonucleotide probes were designed, validated and used to quantify predominant groups of anaerobic bacteria in human faecal samples (Franks et al., 1998). The combination of the two Bacteroides probes for the B. fragilis and B. distasonis groups detected a mean of 5.4 ¥ 1010 cells g- 1 (dry weight) of faeces. The Clostridium coccoides–Eubacterium rectale group-specific probe detected a mean of 7.2 ¥ 1010 cells g- 1 (dry weight) of faeces. Streptococcus–Lactobacillus group-specific probes detected cells ranging in number from 1.7 ¥ 107 to 7 ¥ 108. The future aim of this work is to have a set of about ten probes that can detect more than 90% of the colonic biota in large phylogenetic groups. This approach which combines the power of molecular techniques with modern image analysis at the single cell level will provide much insight into structure–function relationships within gastrointestinal microbial ecosystems.

Conclusions The use of molecular ecology techniques based on nucleic acid probes is likely to revolutionize our approach to microbial ecology in the gastrointestinal tract and will

R.I. Mackie et al.

72

provide, not simply a refinement or increased understanding but a complete description of gastrointestinal community for the first time. Modern molecular ecology techniques based on sequence comparisons of nucleic acids (DNA or RNA) can be used to provide molecular characterization while at the same time providing a classification scheme which predicts natural evolutionary relationships. In principle, nucleic acid probes can be designed to hybridize with a complementary target sequence and thus provide a complete description independent of the growth conditions and media used. An example of the power of these modern molecular ecology techniques is provided by the analysis of SSU rRNA sequences. The highly conserved regions of the SSU rRNA molecule can serve as primer binding sites for in vitro amplification by PCR. The more conserved regions are also useful, serving as targets for universal probes that react with all living organisms or for discriminating between broad phylogenetic groups such as the domains Archaea, Bacteria and Eucarya. The more variable sequence regions are more appropriate for genus, species and sometimes even strain-specific hybridization probes. Thus nucleic acid probes serve to evaluate the presence of specific sequences in the environment and provide a link between knowledge obtained from pure cultures and the microbial populations they represent in the gastrointestinal tract. Furthermore, whole-cell hybridization using in situ PCR is a powerful technique which can be used to describe an organism’s expression of key enzymes. Thus development of these procedures and techniques will result in greater insights into community structure and activity of gut microbial communities in relation to functional interactions between different bacteria, spatial and temporal relationships between different microorganisms and between microorganisms and feed particles. The successful development and application of these methods promises to provide the first opportunity to link distribution and identity of gastrointestinal microbes in their natural environment with their genetic potential and in situ activities.

References Allison, M.J., Hammond, A.C. and Jones, R.J. (1990) Detection of ruminal bacteria that degrade toxic dihydroxypyridine compounds produced from mimosine. Applied and Environmental Microbiology 56, 590–595. Allison, M.J., Mayberry, W.R., McSweeney, C.S. and Stahl, D.A. (1992) Synergistes jonesii, gen. nov., sp. nov.: a rumen bacterium that degrades toxic pyridinediols. Systematic and Applied Microbiology 15, 522–529. Amann, R.I. (1995) Fluorescently labelled, rRNA-targeted probes in the study of molecular ecology. Molecular Ecology 4, 543–554. Amann, R.I. and Kuhl, M. (1998) In situ methods for the assessment of microorganisms and their activities. Current Opinion in Microbiology 1, 352–358. Amann, R.I., Krumholz, L. and Stahl, D.A. (1990) Fluorescent-oligonucleotide probing of whole cells for determinative, phylogenetic, and environmental studies in microbiology. Journal of Bacteriology 172, 762–770. Amann, R.I., Lin, C., Key, R., Montgomery, L. and Stahl, D.A. (1992) Diversity among Fibrobacter isolates: towards a phylogenetic classification. Systematic and Applied Microbiology 15, 23–31. Amann, R.I., Ludwig, W. and Schleifer, K.H. (1994) Identification of uncultured bacteria: a challenging task for molecular taxonomists. ASM News 60, 360–365.

Molecular Ecology and Diversity in Gut Microbial Ecosytems

73

Amann, R.I., Ludwig, W. and Schleifer, K.H. (1995) Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiological Reviews 59, 143–169. Amann, R.I., Snaidr, J., Wagner, M., Ludwig, W. and Schleifer, K.H. (1996) In situ visualization of high genetic diversity in a natural microbial community. Journal of Bacteriology 178, 3496–3500. Attwood, G., Lockington, R.A., Xue, G.P. and Brooker, J.D. (1988) Use of a unique gene sequence as a probe to enumerate a strain of Bacteroides ruminicola introduced into the rumen. Applied and Environmental Microbiology 54, 534–539. Cann, I.K.O., Kocherginskaya, S.A. and White, B.A. (1996) Denaturing gradient gel electrophoresis analysis of polymerase chain-reaction amplified genes coding for 16S rRNAs from ruminal fibrolytic bacteria. Proceedings of Japanese Society of Rumen Metabolism and Physiology 7, 10. Collinson, M.E. and Hooker, J.J. (1991) Fossil evidence of interactions between plants and plant-eating mammals. Philosophical Transactions of the Royal Society, London, Series B 333, 197–200. DeLong, E.F., Wickham, G.S. and Pace, N.R. (1989) Phylogenetic stains: ribosomal RNA-based probes for the identification of microbial cells. Science 243, 1360–1363. Doré, J. and Stahl, D.A. (1991) Phylogeny of anaerobic rumen Chytridiomycete inferred from small subunit ribosomal RNA sequence comparisons. Canadian Journal of Botany 69, 1964–1971. Doré, J., Gramet, G., Goderel, I. and Pochart, P. (1998) Culture independent characterization of human faecal flora using rRNA-targeted hybridization probes. Genetic Selection and Evolution 30(suppl.1), 287–296. Embley, T.M., Finlay, B.J., Dyal, P.L., Hirt, R.P., Wilkinson, W. and Williams, A.G. (1995) Multiple origins of anaerobic ciliates with hydrogenosomes within the radiation of aerobic ciliates. Proceedings of the Royal Society of London 262, 87–93. Farlow, J.O. (1987) Speculations about the diet and digestive physiology of herbivorous dinosaurs. Paleobiology 13, 60–72. Fenchel, T. and Finlay, B.J. (1995) Ecology and Evolution in Anoxic Worlds. Oxford University Press, New York. Flint, H.J., McPherson, C.A., Argustin, G. and Stewart, C.S. (1990) Use of a cellulose-encoding gene probe to reveal restriction fragment length polymorphisms among ruminal strains of Bacteroides succinogenes. Current Microbiology 20, 63–67. Finlay, B.J., Esteban, G., Clarke, K.J., Williams, A.G., Embly, T.M. and Hirt, R.P. (1994) Some rumen ciliates have endosymbiotic methanogens. FEMS Microbiology Letters 117,157–162. Franks, A.H., Harmsen, H.J.M., Raangs, G.C., Jansen, G.J., Schut, F. and Welling, G.W. (1998) Variations of bacterial populations in human feces measured by fluorescent in situ hybridization with group specific 16S rRNA-targeted oligonucleotide probes. Applied and Environmental Microbiology 64, 3336–3345. Gaskins, H.R. (1997) Immunological aspects of host/microbiota interactions at the intestinal epithelium. In: Mackie, R.I., White, B.A. and Isaacson, R.E. (eds) Gastrointestinal Microbiology, Volume 2. Chapman and Hall, New York, pp. 537–587. Guthrie, R.D. (1984) Mosaics, allelochemics, and nutrients: an ecological theory of late Pleistocene megafaunal extinctions. In: Martin, D.S. and Klein, R.G. (eds) Quartenary Extinctions: a Prehistoric Revolution. University Arizona Press, Tucson, pp. 259–298. Head, I.M., Saunders, J.R. and Pickup, R.W. (1998) Microbial evolution, diversity, and ecology: a decade of ribosomal RNA analysis of uncultivated microorganisms. Microbial Ecology 35, 1–21. Hespell, R.B., Akin, D.E. and Dehority, B.A. (1997) Bacteria, fungi and protozoa of the rumen. In: Mackie, R.I., White, B.A. and Isaacson, R. (eds) Gastrointestinal Microbiology, Volume 2. Chapman and Hall, New York, pp. 59–186.

74

R.I. Mackie et al. Hotton, N., Olson, E.C. and Beerbower, R. (1996) Amniote origins and the discovery of herbivory. In: Sumida, S.S. and Martin, K.L.M. (eds) Amniote Origins. Academic Press, New York, pp. 207–264. Hugenholtz, P. and Pace, N.R. (1996) Identifying microbial diversity in the natural environment: a molecular phylogenetic approach. Trends in Biotechnology 14, 190–197. Hume, I.D. and Warner, A.C.I. (1980) Evolution of microbial digestion in mammals. In: Ruckebusch, Y. and Thievend, P. (eds) Digestive Physiology and Metabolism in Ruminants. MTD Press, Lancaster, pp. 665–684. Hungate, R.E. (1976) Microbial activities related to mammalian digestion and absorption of food. In: Spiller, G.A. and Amen, R.J. (eds) Fiber in Human Nutrition. Plenum Press, New York, pp. 131–149. Hungate, R.E. (1984) Microbes of nutritional importance in the alimentary tract. Proceedings of the Nutrition Society 43, 1–11. Hunter-Cevera, J. (1998) The value of microbial diversity. Current Opinion in Microbiology 1, 278–285. Klieve, A. and Swain, R.A. (1993) Estimation of ruminal bacteriophage numbers by pulsed-field gel electrophoresis and laser densitometry. Applied and Environmental Microbiology 59, 2299–2303. Kocherginskaya, S.A., Simpson, J.M. and White, B.A. (1997) Microbial community structure of the rumen as assessed by denaturing gradient gel electrophoresis of polymerase chain-reaction amplified 16S rDNA genes. Evolution of the Rumen Microbial Ecosystem, RRI-INRA Rumen Microbiology Symposium, p. 37. Krause, D.O. and Russell, J.B. (1996) An rRNA approach for assessing the role of obligate amino acid-fermenting bacteria in ruminal amino acid deamination. Applied and Environmental Microbiology 62, 815–821. Krumholz, L.R., Forsberg, C.W. and Veira, D.M. (1983) Association of methanogenic bacteria with ruminal protozoa. Canadian Journal of Microbiology 29, 676–680. Langer, P. (1991) Evolution of the digestive tract in mammals. Verhandlungen der Deutschen Zoologischen Gesellschaft 84, 169–193. Lepper, B.T., Frolking, T.A., Fisher, D.C., Goldstein, G., Sanger, J.E., Wymer, D.A., Ogden, J.G. and Hooge, P.E. (1991) Intestinal contents of a late Pleistocene mastodon from midcontinental North America. Quaternary Research 36, 1–6. Lin, C. and Stahl, D.A. (1995) Taxon-specific probes for the cellulolytic genus Fibrobacter reveal abundant and novel equine-associated populations. Applied and Environmental Microbiology 61, 1348–1351. Lin, C.Z., Raskin, L. and Stahl. D.A. (1997) Microbial community structure in gastrointestinal tracts of domestic animals: Comparative analyses using rRNA-targeted oligonucleotide probes. FEMS Microbiology Ecology 22, 281–294. Ludwig, W., Dorn, S., Springer, N., Kirchof, G. and Schleifer, K.H. (1994) PCR-based preparation of 23S rRNA-targeted group-specific polynucleotide probes. Applied and Environmental Microbiology 60, 3236–3244. Mackie, R.I. (1987) Microbial digestion of forages in herbivores. In: Hacker, J.B. and Ternouth, J.H. (eds) The Nutrition of Herbivores. Academic Press, Sydney, pp. 233–265. Mackie, R.I. and Doré, J. (1997) Molecular ecology of gastrointestinal ecosystems. In: Martins, M.T., Sato, M.I.Z., Tiedje, J.M., Hagler, L.C.N., Dobereiner, J. and Sanchez, P.S. (eds) Progress in Microbial Ecology. SBM/ICOME, Sao Paulo, Brazil, pp. 317–328. Mackie, R.I., Sghir, A. and Gaskins, H.R. (1999) Developmental microbial ecology of the neonatal gastrointestinal tract. American Journal of Clinical Nutrition 69 (suppl), 1035S1045S. McSweeney, C.S. and Mackie, R.I. (1997) Gastrointestinal detoxification and digestive disorders in ruminant animals. In: Mackie, R.I. and White, B.A. (eds) Gastrointestinal Microbiology, Vol. 1. Chapman and Hall, New York, pp. 583–634.

Molecular Ecology and Diversity in Gut Microbial Ecosytems

75

McSweeney, C.S., Mackie, R.I., Odenyo, A.A. and Stahl, D.A. (1993) Development of an oligonucleotide probe and its application for detection and quantitation of the ruminal bacterium Synergistes jonesii in a mixed-population chemostat. Applied and Environmental Microbiology 59, 1607–1612. Mannarelli, B., Stack, R.J., Lee, D. and Ericsson, L. (1990) Taxonomic relatedness of Butyrivibrio, Lachnospira, Roseburia, and Eubacterium species as determined by DNA hybridization and extracellular-polysaccharide analysis. International Journal of Systematic Bacteriology 40, 370–378. Muyzer, G. and Smalla, K. (1998) Application of denaturing gradient gel electrophoresis (DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial ecology. Antonie van Leeuwenhoek Journal of Microbiology 73, 27–141. Muyzer, G., Brinkoff, T., Nubel, U., Santegoeds, C., Schafer, H. and Wawer, C. (1998) Denaturing gradient gel electrophoresis (DGGE) in microbial ecology. Molecular Microbial Ecology Manual 3.4.4:1–27. Kluwer Academic Publishers, The Netherlands. Muyzer, G., de Waal, E.C. and Uitterlinden, A.G. (1993) Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reactionamplified genes coding for 16S rRNA. Applied Environmental Microbiology 59, 695–700. Ning, Z., Attwood, G.T., Lockington, R.A. and Brooker, J.D. (1991) Genetic diversity in ruminal isolates of Selenomonas ruminantium. Current Microbiology 22, 279–284. Odenyo, A.A., Mackie, R.I., Stahl, D.A. and White, B.A. (1994a) The use of 16S rRNA-targeted oligonucleotide probes to study competition between ruminal fibrolytic bacteria: development of probes for Ruminococcus species and evidence for bacteriocin production. Applied and Environmental Microbiology 60, 3688–3696. Odenyo, A.A., Mackie, R.I., Stahl, D.A. and White, B.A. (1994b) The use of 16S rRNA-targeted oligonucleotide probes to study competition between ruminal fibrolytic bacteria: pure-culture studies with cellulose and alkaline peroxide-treated wheat straw. Applied and Environmental Microbiology 60, 3697–3703. Ogata, K., Aminov, R.I., Nagamine, T., Sugiura, M., Tajima, K., Mitsumori, M., Sekizaki, T., Kudo, H., Minato, H. and Benno, Y. (1997) Construction of a Fibrobacter succinogenes genomic map and demonstration of diversity at the genomic level. Current Microbiology 35, 22–27. Okhuma, M. and Kudo, T. (1996) Phylogenetic diversity of the intestinal bacterial community in the termite Reticulotermes speratus. Applied and Environmental Microbiology 62, 461–468. Orpin, C.G. and Joblin, K.N. (1997) The rumen anaerobic fungi. In: Hobson, P.N. and Stewart, C.S. (eds) The Rumen Microbial Ecosystem, 2nd edn. Chapman and Hall, New York, pp. 140–195. Pace, N.R. (1997) A molecular view of microbial diversity and the biosphere. Science 276, 734–740. Pace, N.R., Stahl, D.A., Lane, D.J. and Olsen, G.J. (1986) The analysis of natural microbial populations by ribosomal RNA sequences. Advances in Microbial Ecology 9, 1–55. Poulsen, L.K., Lon, F., Kristensen, C.S., Hobolth, P., Molin, S. and Krogfelt, K.A. (1994) Spatial distribution of Escherichia coli in the mouse large intestine inferred from rRNA in situ hybridization. Infection and Immunity 62, 5191–5194. Poulsen, L.K., Licht, T.R., Rang, C., Krogfelt, K.A. and Molin, S. (1995) The physiological state of E. coli growing in the large intestine of streptomycin-treated mice. Journal of Bacteriology 177, 5840–5845. Raskin, L., Capman, W.C., Sharp, R., Poulson, L.K. and Stahl, D.A. (1997) Molecular ecology of gastrointestinal systems. In: Mackie, R.I., White, B.A. and Isaacson, R. (eds) Gastrointestinal Microbiology, Vol. 2. Chapman and Hall, New York, pp. 243–298. Rhodes, A.N., Urbance, J.W., Youga, H., Corlew-Newman, H., Reddy, C.A., Klug, M.J., Tiedje,

76

R.I. Mackie et al. J.M. and Fisher, D.C. (1998) Identification of bacterial isolates obtained from intestinal contents associated with 12,000-year-old mastodon remains. Applied and Environmental Microbiology 64, 651–658. Rolfe, R.D. (1997) Colonization resistance. In: Mackie, R.I., White, B.A. and Isaacson, R.E. (eds) Gastrointestinal Microbiology, Vol. 2. Chapman and Hall, New York, pp. 501–536. Rollo, F. and Marota, I. (1999) How microbial ancient DNA, found in association with human remains, can be interpreted. Philosophical Transactions of the Royal Society, London Series B 354, 111–119. Sayler, G.S. and Layton, A.C. (1990) Environmental application of nucleic acid hybridization. Annual Review of Microbiology 44, 625–648. Schopf, J.W. (1993) Microfossils of the early archaean apex chert: new evidence of the antiquity of life. Science 260, 640–646. Sharp, R., Zeimer, C.J., Stern, M.D. and Stahl, D.A. (1998) Taxon-specific associations between protozoal and methanogen populations in the rumen and a model rumen system. FEMS Microbiology Ecology 26, 71–78. Shi, Y. and Weimer, P.J. (1997) Competition for cellobiose among three predominant ruminal cellulolytic bacteria under substrate-excess and substrate-limited conditions. Applied and Environmental Microbiology 63, 743–748. Shi, Y., Odt, C.L. and Weimer, P.J. (1997) Competition for cellulose among three predominant ruminal cellulolytic bacteria under substrate-excess and substrate-limited conditions. Applied and Environmental Microbiology 63, 734–742. Simpson, J.M., McCracken, V.J., White, B.A., Gaskins, H.R. and Mackie, R.I. (1999) Application of denaturant gradient gel electrophoresis for the analysis of the porcine gastrointestinal microbiota. Journal of Microbiological Methods 36, 167–179. Stackebrandt, E. (1997) The biodiversity convention and its consequences for the inventory of prokaryotes. In: Martins, M.T., Sato, M.I.Z., Tiedje, J.M., Hagler, L.C.N., Dobereiner, J. and Sanchez, P.S. (eds) Progress in Microbial Ecology. SBM/ICOME, Sao Paulo, Brazil. Stahl, D.A. (1993a) The natural history of microorganisms. ASM News 59, 609–613. Stahl, D.A. (1993b) Comparison of nucleic acids from microorganisms: sequencing approaches. Molecular evolution: producing the biochemical data. Methods in Enzymology 224, 373–391. Stahl, D.A. (1997) Molecular approaches for the measurement of density, diversity and phylogeny. In: McInerney, M.J., Walter, M.V., Stetzenback, L.D. and Walter, M.V. (eds) Manual of Environmental Microbiology. ASM Press, Washington, DC, pp. 102–114. Stahl, D.A. and Amann, R. (1991) Development and application of nucleic acid probes. In: Stackebrandt, E. and Goodfellow, M. (eds) Nucleic Acid Techniques in Bacterial Systematics. John Wiley & Sons, Chichester, UK, pp. 205–248. Stahl, D.A., Flesher, B., Mansfield, H.R. and Montgomery, L. (1988) Use of phylogenetically based hybridization probes for studies of ruminal microbial ecology. Applied and Environmental Microbiology 54, 1079–1084. Stewart, C.S., Flint, H.J. and Bryant, M.P. (1997) The rumen bacteria. In: Hobson, P.N. and Stewart, C.S. (eds) The Rumen Microbial Ecosystem, 2nd edn. Chapman and Hall, New York, pp. 10–72. Stumm, C.K., Gizjen, H.J. and Vogels, G.D. (1982) Association of methanogenic bacteria with ovine rumen ciliates. British Journal of Nutrition 47, 95–99. Sues, H.-D. and Reisz, R.R. (1998) Origins and early evolution of herbivory in tetrapods. Trends in Ecology and Evolution 13, 141–145. Tajima, K., Aminov, R.I., Nagamine, T., Ogata, K., Nakamura, M., Matsui, H. and Benno, Y. (1999) Molecular bacterial diversity of the rumen as determined by sequence analysis of 16S rDNA libraries. FEMS Microbiology Ecology 27, 159–169. Tannock, G.W. (1997) Influences of the normal microbiota on the animal host. In: Mackie, R.I.,

Molecular Ecology and Diversity in Gut Microbial Ecosytems

77

White, B.A. and Isaacson, R.E. (eds) Gastrointestinal Microbiology, Vol. 2. Chapman and Hall, New York, pp. 466–497. Thill, L.A., Kocherginskaya, S.A., Naylor, G.E., Joblin, K.N. and White, B.A. (1997) Small subunit ribosomal RNA sequence analysis of anaerobic ruminal fungi isolated in New Zealand. Proceedings of the 24th Biennial Conference on Rumen Function, p. 28. Thomasson, J.R. and Voorhies, M.R. (1990) Grasslands and grazers. In: Briggs, D.E.G. and Crowther, P.R. (eds) Paleobiology: a Synthesis. Blackwell Scientific Publishers, Oxford, pp. 84–87. Vogels, G.D., Hoppe, W.F. and Wheelis, M.L. (1980) Association of methanogenic bacteria with rumen ciliates. Applied and Environmental Microbiology 40, 608–612. Ward, D.M. (1989) Molecular probes for analysis of microbial communities. In: Characklis, W.G. and Wilderer, P.A. (eds) Structure and Function of Biofilms. John Wiley & Sons, Inc., New York, pp. 145–163. Ward, D.M., Bateson, M.M., Weller, R. and Ruff-Roberts, A.L. (1992) Ribosomal RNA analysis of microorganisms as they occur in nature. Advances in Microbial Ecology 12, 219–286. Whitehead, T.R. and Cotta, M.A. (1993) Development of a DNA probe for Streptococcus bovis by using a cloned amylase gene. Journal of Clinical Microbiology 31, 2387–2391. Whitford, M.F., Forster, R.J., Beard, C.E., Gong, J. and Teather, R.M. (1998) Phylogenetic analysis of rumen bacteria by comparative sequence analysis of cloned 16S rRNA genes. Anaerobe 4, 153–163. Whitman, W.B., Coleman, D.C. and Wiebe, W.J. (1998) Prokaryotes: the unseen majority. Proceedings of the National Academy of Sciences USA 95, 6578–6583. Williams, A.G. and Coleman, G.S. (1997) The rumen protozoa. In: Hobson, P.N. and Stewart, C.S. (eds) The Rumen Microbial Ecosystem, 2nd edn. Chapman and Hall, New York, pp. 73–139. Wilson, K.H. and Blitchington, R.B. (1996) Human colonic biota studied by ribosomal DNA sequence analysis. Applied and Environmental Microbiology 62, 2273–2278. Wintzingerode, F.V., Gobel, U.B. and Stackebrandt, E. (1997) Determination of microbial diversity in environmental samples: pitfalls of PCR-based rRNA analysis. FEMS Microbiology Reviews 21, 213–229. Woese, C.R. (1994) There must be a prokaryote somewhere: microbiology’s search for itself. Microbiology Reviews 58, 1–9. Woese, C.R., Kandler, O. and Wheelis, M.L. (1990) Towards a natural system of organisms: proposal for the domains Archaea, Bacteria, and Eucarya. Proceedings of the National Academy of Sciences USA 87, 4576–4579. Wright, A.-D.G. and Lynn, D.H. (1997a) Phylogenetic analysis of the rumen ciliate family Ophryoscolecidae based on 18S ribosomal RNA sequences, with new sequences from Diplodinium, Eudiplodinium, and Ophryoscolex. Canadian Journal of Zoology 75, 963–970. Wright, A.-D.G. and Lynn, D.H. (1997b) Monophyly of the Trichostome ciliates (phylum Ciliophora: class Litostomatea) tested using new 18S rRNA sequences from the Vestibuliferids, Isotrichia intestinalis and Dasytricha ruminantium, and the Haptorian, Didinium nasutum. European Journal of Protistology 33, 305–315. Wright, A.-D.G., Dehority, B.A. and Lynn, D.H. (1997) Phylogeny of the rumen ciliates Entodinium, Epidinium, and Polyplastron (Litostomatea: Entodiniomorphida) inferred from small subunit ribosomal RNA sequences. Journal of Eukaryotic Microbiology 44, 61–67. Zoetendahl, E.G., Akkermans, A.D.L. and de Vos, W.M. (1998) Temperature gradient gel electophoresis analysis of 16S rRNA from human fecal samples reveals stable and host-specific communities of active bacteria. Applied and Environmental Microbiology 64, 3854–3859.

5

Microbial Adherence to the Plant Cell Wall and Enzymatic Hydrolysis C.W. FORSBERG1, E. FORANO2 AND A. CHESSON3 1Department

of Microbiology, University of Guelph, Guelph, Ontario, Canada; 2Laboratoire de Microbiologie, INRA CR de Clermont-Ferrand–Theix, Saint-Genès-Champanelle, France; 3Rowett Research Institute, Bucksburn, Aberdeen, UK

Interactions of ruminal microorganisms with plant cell walls The ruminal environment is a highly complex ecosystem in which the plant cell wall digesting microorganisms constitute a minor proportion of the total population, but make available the carbon and energy for the ecosystem when animals are fed a forage diet. The major cellulolytic bacteria, Ruminococcus albus, Ruminococcus flavefaciens and Fibrobacter succinogenes, may account for 0.3–4% of the bacterial population (Krause et al., 1999; Weimer et al., 1999). The fungi may account for 8% of the microbial biomass (Orpin, 1983), and only a portion of these produce highly active cellulases and hemicellulases (Trinci et al., 1994). A limited number of protozoal genera as well have an important role in digesting plant cell walls, and they may digest 5–21% of the cellulosic materials dependent upon the diet (Dijkstra and Tamminga, 1995). The primary factor limiting cellulose digestion appears to be the availability of cellulose digestion sites in the plant material rather than low cellulolytic activity (Dehority and Tirabasso, 1998). This conclusion is supported by similar studies by Weimer et al. (1999). However, Weimer et al. (1999) reported that differences in cellulolytic populations in individual cows were larger than those attributable to diet, suggesting that each cow maintains a unique assemblage of cellulolytic species. This could result from differing efficiencies in masticating plant cell walls since the bulk of microbial digestion occurs in the secondary cell wall (Wilson and Mertens, 1995). The fascinating aspect of these studies is that they present a general overview of population dynamics in vivo which begs the question of the factors that regulate adhesion of microorganisms to fibrous substrates, the mechanism of hydrolysis, and the cellular and extracellular factors influencing degradation of individual but highly integrated polymers making up the plant cell wall.

© CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

79

80

C.W. Forsberg et al.

The ‘cellulosome’ paradigm and the molecular basis for adhesion Many studies have been undertaken to elucidate the molecular basis of adhesion and to further improve adhesion abilities of rumen microorganisms to the plant particles. Researchers have focused on the isolation of cellulose-binding proteins (CBPs) in order to examine their role in adhesion of microorganisms, and have searched for dockerin domains on polysaccharidase sequences, and scaffoldins in order to identify enzymatic complexes similar to cellulosomes. To complement this research, mutant strains exhibiting impaired adhesion to cellulose have been isolated.

The cellulosome paradigm The cellulosome is a multi-functional, multi-enzyme complex of high molecular mass, which is attached to the cell surface and mediates both cellular adhesion to cellulose and efficient solubilization of crystalline cellulose. The cellulosome structure was first discovered and described in the anaerobic thermophilic bacterium Clostridium thermocellum, and is considered today as one of the major molecular paradigms of bacterial cellulolysis and adhesion to cellulose (Bayer et al., 1998). Biochemical and ultrastructural staining techniques showed that this complex (about 2 MDa) aggregates into polycellulosomes (up to 100 MDa) that form protuberances on the cell surface. The C. thermocellum cellulosome was shown to be composed of a 250 kDa non-catalytic glycoprotein, named CipA, that binds to cellulose, and of a variety of glycosyl hydrolases (endoglucanase, cellobiohydrolase, xylanase or lichenase). CipA was further shown to act as a scaffoldin protein that anchors the cellulosomal enzymes through receptor domains (Fig. 5.1). Indeed, all of the enzymes that assemble in a cellulosome structure have at their C-terminus, a highly conserved, non-catalytic domain called a dockerin, which is composed of two similar segments of 22 amino acids each (Bayer et al., 1998) that specifically bind to nine receptor domains called cohesins present in the CipA scaffold structure (Fig. 5.1). CipA also contains a cellulose-binding domain (CBD) mediating binding of the entire cellulosome to cellulose, and a dockerin domain (type II) that tethers the cellulosome to the cell through another type of cohesin–dockerin interaction and a complex set of proteins (Beguin and Lemaire, 1996; Bayer et al., 1998). Complexes similar to C. thermocellum cellulosomes have been described for mesophilic cellulolytic clostridia such as Clostridium cellulovorans, Clostridium cellulolyticum, Clostridium josui, Clostridium papyrosolvens; the non-clostridial anaerobic bacteria Bacteroides cellulosolvens and Acetivibrio cellulolyticus; and also for the aerobic bacterium Vibrio sp. (Bayer et al., 1998). The evidence in favour of cellulosomes in these bacteria include: (i) the production of high molecular weight complexes; (ii) the presence of cell-surface protuberances; (iii) cross-reactivity with antibodies against the C. thermocellum cellulosome; and (iv) isolation and sequencing of genes encoding the scaffoldin and cellulosomal enzymes. The various scaffoldins may differ in the number of cohesins, the position of the CBD (internal or terminal) and the organization of the dockerin on the polypeptide. The cellulosomal structure seems widespread and thus would seem to have been selected through evolution as a very efficient cellulolytic system.

Microbial Adherence to the Plant Cell Wall

81

Adhesion to cellulose

1

2

CBD

3

4

5

6

7

8

9

Cohesion domain (type I)

Cellulose binding domain

Dockerin domain (type I)

Catalytic domain

Dockerin domain (type II)

Linker

Anchoring to the cell

Fig. 5.1. Schematic representation of the cellulosome of Clostridium thermocellum.

The cellulosome equivalent in rumen microorganisms Three types of evidence indicate the presence of cellulosome-like structures in rumen cellulolytic microorganisms. First, protuberances similar to those found on C. thermocellum were observed on R. albus, R. flavefaciens and F. succinogenes using electron microscopy and staining with cationized ferritin (Leatherwood, 1973; Lamed et al., 1987; Miron et al., 1989). However, it was recently shown that cationized ferritin stains protuberances on the cell surface of a non-cellulolytic Clostridium species, and that the presence of these protuberances was correlated with growth rate rather than with the presence of cellulosomes (Blair and Anderson, 1998). Consequently, protuberances are not a reliable indicator of cellulosomes. Second, high molecular mass complexes with cellulolytic and/or xylanolytic activity were shown in rumen bacteria and fungi. R. albus 8 was shown to produce large multi-enzyme complexes at the cell surface when cultivated in the presence of rumen fluid or phenylpropanoic acid (Stack and Hungate, 1984); similarly, R. albus SY3 was found to elaborate high molecular mass enzyme aggregates (Wood et al., 1982). More recently, a cellulase complex of 2 MDa was isolated from the culture supernatant of R. albus F40, and was shown to include about ten proteins with endoglucanase or xylanase activity together with a protein of 250 kDa with no catalytic activity that might be a scaffoldin (Ohara et al., 1998). R. flavefaciens FD1 and Butyrivibrio fibrisolvens H17C also produce multi-enzyme aggregates, composed either of several endoglucanases and xylanases, or only of several endoglucanases (Doerner and White, 1990; Lin and Thompson, 1991). The R. flavefaciens complexes were able to degrade crystalline cellulose (Doerner and White, 1990). High molecular mass complexes (4 MDa) with endoglucanase activity were also isolated from membrane fractions and culture fluid of F. succinogenes S85 cultures on cellulose, but they were not able to solubilize crystalline cellulose (Groleau and Forsberg, 1983). The ability of these bacterial cellulase complexes to bind to cellulose was not examined.

82

C.W. Forsberg et al.

Neocallimastix frontalis produces a multi-enzyme complex (750–1000 kDa) that binds to and hydrolyses cellulose, and includes endoglucanase and exoglucanase (in majority) activities (Wood et al., 1995). Similarly, 80% of the cellulase, xylanase and mannanase activity of Piromyces are extracellular and associated with a complex that binds to cellulose (Ali et al., 1995). A multi-enzyme complex of at least ten endoglucanases was also found in the strain E2 of Piromyces (Dijkerman et al., 1997). Third, reiterated sequences resembling dockerins of the clostridial cellulosomal enzymes were found in the great majority of the sequences of polysaccharidases from the rumen fungi and in some of the Ruminococci enzymes (Table 5.1). These sequences are longer than the reiterated dockerin sequences of C. thermocellum or related clostridia, are sometimes internal, and may be found in one to three copies (Table 5.1). The reiterated sequences of the fungi Neocallimastix, Piromyces and Orpinomyces show strong homology to each other. Antibodies raised against the reiterated sequence of xylanase XYNA from Orpinomyces spp. reacted with numerous extracellular proteins of this fungus and also from N. frontalis, suggesting such sequences are widespread and share common epitopes (Li et al., 1997a). Furthermore, the reiterated sequences (fused to glutathione-S-transferase) were shown to bind specifically to proteins of 116 and 97 kDa in Neocallimastix patriciarum and Piromyces, respectively, suggesting that they act as dockerins (Fanutti et al., 1995). A dockerin-like sequence was found in one cellulase from R. albus (Karita et al., 1997) while dockerin-like sequences were found in at least three glycanase enzymes from R. flavefaciens 17 (Kirby et al., 1997). More recently, Ding et al. (1999) identified a 350 kDa glycosylated cellulose-binding protein in R. flavefaciens 17 that was a strong candidate for a scaffoldin/cellulosome-like integrating protein (CIP). The protein was subjected to trypsin digestion and fragments sequenced by Edman degradation. PCR primers designed from the amino acid sequences were used to clone fragments of the encoding gene. Sequencing of the recovered clone resulted in the identification of a series of cohesin-like repeats of 140 residues suggesting that a scaffoldin had indeed been cloned. The Ruminococcus reiterated dockerin sequences showed a significant, although distant, relationship with the dockerins of clostridia (Kirby et al., 1997; Karita et al., 1997); however, they did not show any sequence similarity with those from ruminal fungi. C-terminal reiterated sequences of 164 and 156 amino acids in the CelA endoglucanase of Orpinomyces joyonii exhibited notable similarity to an N-terminal sequence of the EG3 enzyme from F. succinogenes and Fibrobacter intestinalis, but not with enzymes from other rumen organisms. The significance of this observation should become apparent with further studies.

Isolation of cellulose-binding proteins from rumen bacteria CBPs were identified in the three main cellulolytic rumen bacteria, F. succinogenes, R. flavefaciens and R. albus (Mitsumori and Minato, 1995; Pegden et al., 1998), and in the other fibrolytic species, but also in non-cellulolytic species such as Prevotella ruminicola, Megasphaera elsdenii and Eubacterium ruminantium that do not attach to cellulose (Mitsumori and Minato, 1995). Several CBPs ranging from 36 to 240 kDa were found in F. succinogenes S85 and in F. intestinalis DR7 (Gong et al., 1996; Miron and Forsberg, 1998), and some of the

Gene product

Descriptiona

Modular arrangementb

Butyrivibrio fibrisolvens E14 B. fibrisolvens E14 Epidium caudatum Fibrobacter succinogenes S85 F. succinogenes S85 F. succinogenes S85 F. succinogenes BL21 F. succinogenes S85 F. succinogenes S85 F. succinogenes S85 Neocallimastix frontalis MCH3 N. frontalis

CinA CinB Epi3 CedA CelG EGB EGC EGD EGE EGF CelA XYN3

Cinnamoyl ester hydrolase Cinnamoyl ester hydrolase Endoglucanase Cellobiosidase Glucanase Endoglucanase Endoglucanase Endoglucanase Endoglucanase Endoglucanase Endoglucanase Xylanase

Neocallimastix patriciarum N. patriciarum N. patriciarum N. patriciarum N. patriciarum N. patriciarum

BnaA BnaB BnaC CelA CelB XynA

AXE AXE AXE Cellobiohydrolase Glucanase Xylanase

N. patriciarum Orpinomyces joyonii 26 O. joyonii SG4 O. joyonii SG4 Orpinomyces PC-2 Orpinomyces PC-2 Orpinomyces PC-2 Orpinomyces PC-2 Orpinomyces PC-2

XynB CelA CelB 2 CelB 29 AxeA CelA CelB CelC CelE

Xylanase Endoglucanase Cellulase Cellulase AXE Endoglucanase Endoglucanase Endoglucanase Endoglucanase

CD CD CD5 CD5 CD 5A/SL/BTD ?/CD9 CD9/BTD CD9/SL/BTD CD9/BTD CBD?/CD? CD5/EL/TR2–38–36 CD11/SL/CD11/TL/TR2 ⫺40,40 CD/TL/TR2–37,39 ?/CD CD/TR CBD/NL/CD6 CD5/STL/TR2–36,37 CD11/SL/CD11/TL/TR2 ⫺40,40 CD10/L/TR12–8 CD5/TR2–164,156 CD5/XL/Tdoc1–38 CD5/XL/TR2–36,37 CD TR2–26,34/PL/CD6 CD5/L/TR2–36,37 TR2–26,29/TL/CD6 CD10/L/TR2–35–37

Amino acid residues Km/Vmaxc 246 285 341 357 519 555 620 668 467 1053 ? 607 393 392 389 428 473 607 860 910 585 513 313 459 471 449 477

⫺/76*

⫺/16.5*

⫺/11.5* ⫺/57*

0.24/⫺/5,980

0.9/785 ⫺/9.7* 14.5*

Gen Bank # U44893 U44893 AB011273 U07419 U33887 L14436 L48039 U05897 U05897 U39070 U38843 X82266 U66251 U66252 U66253 U29872 Z31364 X65526 S71569 U59432 AF015249 AF015248 AF001178 U63837 U57818 U63838 U97153 (Continued)

83

Organism

Microbial Adherence to the Plant Cell Wall

Table 5.1. Recently characterized glycoside hydrolases and esterases from ruminal bacteria, fungi and protozoa.

84

Table 5.1. Continued

Organism

Gene product

Descriptiona

Modular arrangementb

Orpinomyces PC-2 Orpinomyces PC-2 Prevotella ruminicola B14 P. ruminicola B14 P. ruminicola B14 P. ruminicola B14 P. ruminicola 23 Piromyces spp. Piromyces spp. Piromyces spp. Piromyces spp. Polyplastron multivesiculatum P. multivesiculatum Ruminococcus albus 7 R. albus F-40 Ruminococcus flavefaciens 17

LICA XynA CdxA XynA XynB XynC D31d MANA MANB MANC XYLA PolyX XYNA XynA EGIV XynB

Lichenase Xylanase b -Glucosidase Xylanase Glycosidase Xylanase Xylanase Mannanase Mannanase Mannanase Xylanase Xylanase Xylanase Xylanase Endoglucanase Xylanase/AXE

R. flavefaciens 17

XynD

R. flavefaciens 17 R. flavafaciens FD-1 Streptococcus bovis JB1

EndA CelD

Xylanase/ b glucanase Endoglucanase Endoglucanase b -(1,3–1,4) glucanase

CD16 CD11/L/TR2–40,40 CD3 CD10 CD ? CD10/?/CD10 CD10/TR/CD10 CD26/TR3–36,36,36 CD26/NL/TR2–36,37 CD26/NL/TR2–36,37 CD11/TR2–39,36/CD11 CD11 CD11 CD CD5/TR2–32,31 CD11/TSD/CDAXE/TL/ TR 2–32,33 CD11/TSD/TR232,33/TL/CD16 CD5/?/TL/TL/TR2–32 CD9 CD16

245 362 789 369 319 560

0.91/5,320

Gen Bank #

623 175 218 680 312 781

U63813 U57819 U35425 Z49241 Z49241 Z79595 U53926 X91858 X97408 X97520 X91857 AB011274 AJ009828 U43089 D16315 Z35226

802

S61204

759 405 237

Z83304 L05368 Z92911 1

5.0/3.5

606

2.8/338

AXE, acetyl xylan esterase. Modular arrangement: CD#, catalytic domain followed by the number of the glycosyl family or a ? if there is no similarity to a known family; CBD, cellulose binding domain; BTD, basic terminal domain; L, GL, SL, TL, STL, PL, Linker regions with designation, or rich in glutamic acid, serine, threonine or proline; TR, dockerin with a tandem repeat, the first number indicates the number of repeats and the subsequent numbers indicate the amino acid residues in each; Tdoc, dockerin in a single sequence; X, not known; TSD, thermostabilizing domain. c The number before the slash is the K (mM) and the number after it is the V ⫺1 mg⫺1 protein) where values are available. Where an asterisk is shown, m max (µmol min the specific activity was listed using the relevant substrate. b

C.W. Forsberg et al.

a

Amino acid residues Km/Vmaxc

Microbial Adherence to the Plant Cell Wall

85

proteins of 120 kDa and lower had endoglucanase activity. Two CBPs of 120 and 225 kDa were purified from F. succinogenes S85, and the gene coding for the 120 kDa protein was cloned in Escherichia coli and sequenced (Mitsumori et al., 1996). This protein was later shown to be an endoglucanase bearing a CBD (Malburg et al., 1997). In parallel, Gong et al. (1996) recognized two other CBPs, of 180 and 240 kDa. The 180 kDa protein is a glycosylated xylanase (E.E. Egbosimba and C.W. Forsberg, unpublished data) that shares common epitopes with numerous proteins from the outermembrane. Furthermore, polyclonal antibodies raised against this CBP decreased the adhesion of F. succinogenes to cellulose by 60%, suggesting that it and related proteins have an important role in adhesion. To identify proteins important in adhesion of F. intestinalis to cellulose, polyclonal antibodies prepared against whole cells were adsorbed with cells of the non-adherent mutant (Miron and Forsberg, 1998, Section 2.4) to remove all antibodies reacting with surface epitopes not involved in the adhesion process (Miron and Forsberg, 1999). These adsorbed antibodies reacted strongly with six of the major CBPs of strain DR7, but reacted very weakly with a non-adherent mutant. The antiserum was shown to react with glycosyl residues rather than protein. Since the non-adherent mutant was not missing any of the CBPs, glycosylation appears to have an important role in the adhesion process. Monosaccharide analysis of the CBPs showed that they contained mainly galactosamine, glucosamine, galacturonic acid and glucuronic acid. Of these compounds, glucosamine and galacturonic acid, each at 10%, blocked binding. These data were interpreted to indicate that the compounds interfered with binding by occupying adhesion sites on the cellulose substrate; however, the precise role of the glycosyl components of the CBPs involved in adhesion is not known. Recently, two low-molecular mass cellulose-binding polypeptides (16 and 21 kDa) were isolated from R. albus 8 (Pegden et al., 1998). One of them (CbpC) possessed structural motifs typical of the Pil protein family, which is comprised mostly of type 4 fimbrial proteins produced by Gram-negative pathogenic bacteria. Inclusion of either ruminal fluid or phenyl propionic acid in the growth medium, a treatment that was shown to increase R. albus adherence to cellulose, also increased the concentration of CbpC transcripts (Pegden et al., 1998). DNA sequences homologous to the CbpC gene were found in other strains of R. albus. The authors suggest that fimbrial-type adhesion proteins may represent a novel strategy (in addition to the cellulosome?) for the adhesion of Gram-positive bacteria to cellulose.

Isolation of non-adherent mutants Adherence deficient mutants of F. succinogenes, F. intestinalis and R. albus have been obtained (Gong and Forsberg, 1989; Miron and Forsberg, 1998; Miron et al., 1998). The mutants were spontaneous and obtained by enrichment of the non-adherent fraction of bacterial cultures, and the genes involved in the mutations have not been identified. Mutagenesis with transposons, which allows easy identification of the mutated gene(s), was also applied to F. succinogenes and R. albus, but was not successful (Mosoni, 1999). An adhesion-defective mutant of F. intestinalis DR7, isolated by Miron and Forsberg (1998), showed limited growth on insoluble cellulosic substrates. The mutant

86

C.W. Forsberg et al.

displayed lower cellulose-degrading activities than the wild type, and these activities were located in the extracellular fluid, while they were mostly cell-bound in the wild type. This may suggest that the mutant is impaired in assembling multi-enzyme complexes involved in both adhesion and hydrolysis of cellulose. Four mutants isolated from the strain S85 of F. succinogenes displayed different phenotypes regarding their capability to utilize cellulose (Gong and Forsberg, 1989), suggesting that the adhesion mechanism may require several factors. An adhesion-defective derivative isolated from R. albus SY3 was also impaired in its capacity to grow on cellulose (Miron et al., 1998). Two adhesion-defective mutants were also isolated from R. albus 20; they showed both a decreased rate and reduced extent of cellulose degradation (Mosoni, 1999). In addition to these studies with mutant strains, a great heterogeneity in adhesion and cellulolytic performances was observed among different wild strains of R. albus (Morris and Cole, 1987). Altogether, these results suggest that adhesion of R. albus to cellulose is a complex process that may be mediated by several factors. In conclusion, there is evidence that the rumen fungi and the ruminococci produce multi-enzyme complexes that may reorganize to form cellulosomes. These complexes are likely to be involved in both adhesion of the cell to cellulose and degradation of this substrate. Indeed, for all the fibrolytic microorganisms, adhesion and cellulolytic activity appear closely, although not always strictly, related. However, there is no evidence at this time for a cellulosome-type organization for F. succinogenes. Complete understanding of the molecular basis of adhesion of the main rumen fibrolytic species will rely on further work, such as isolation of genes encoding the putative scaffoldins, and the molecular characterization of the isolated mutants.

Diversity and catalytic properties of bacterial, fungal and protozoan glycoside hydrolases In excess of 100 plant cell wall degrading enzymes have been cloned from ruminal organisms. Recently cloned genes are listed in Table 5.1 while previously cloned genes and their characterized native gene products are described in several reviews (Chesson and Forsberg, 1997; Forsberg et al., 1997). Of great interest is the recent cloning of cellulase and xylanase genes from the ruminal protozoa Epidinium caudatum and Polyplastron multivesiculatum (Devillard et al., 1999; Takenaka et al., 1999). Three glucanases, one cellobiosidase and two xylanases were cloned from E. caudatum while one glucanase, one cellobiosidase and several xylanases were cloned from P. multivesiculatum. Thus we can say that all major rumen fibre-digesting organisms contain multiple hemicellulase and glucanase genes which code for related or different enzymes that can be separated into glycosyl hydrolase families based on the amino acid sequence of the catalytic domain (CD). A current overview of the glycoside hydrolase families (Henrissat, 1998) can be found at the URL: http://afmb.cnrs-mrs.fr/~pedro/CAZY/. Although no crystal structure of a ruminal glycoside hydrolase has been reported, structures of cellulases from non-rumen organisms are available for more than half of the families. Enzymatic hydrolysis of the glycosidic bond takes place via general acid catalysis that involves a pair of carboxylic acids (Glu or Asp) in essentially all cases investigated (McIntosh et al., 1996). This hydrolysis occurs via two major mechanisms giving rise to either retention or inversion of the anomeric configuration (Henrissat et al., 1998).

Microbial Adherence to the Plant Cell Wall

87

In the past, considerable emphasis was placed on characterization of the catalytic properties of glycosidases; however, recently the focus has been on establishing the cellulosome-like structure in ruminal fungi and bacteria. As documented earlier (see pp. 80–82), most enzymes are an integral part of the cellulosome-like structure, but some are not. These include the Orpinomyces PC-2 LICA lichenase and AxeA acetyl xylan esterase, the N. patriciarum CelA cellobiohydrolase and the N. patriciarum BnaB acetyl xylan esterase (Table 5.1). The CelA cellobiohydrolase CD, in contrast to other fungal glycoside hydrolases, exhibited no similarity to those of other ruminal organisms and instead resembled CBHII of Trichoderma reesei. The CelA CBD resembled the type 1 family that is exclusive to aerobic fungal cellulases (Tomme et al., 1995). Knowledge of the mechanism by which this enzyme interacts with the cellulosome-like complex in cellulose hydrolysis will be important to our understanding of cellulolysis. The number and organization of CDs in enzymes of ruminal microorganisms is to a degree characteristic of the genus. Neocallimastix and Piromyces species often have two CDs in a single protein (Gilbert et al., 1992; Fanutti et al., 1995), while Xue et al. (1992) reported three in a broad spectrum glucanase of N. patriciarum. However, Fanutti et al. (1995) reported that the second xylanase gene did not code for a catalytically active protein. In the bacteria, the XynB and XynD xylanases from R. flavefaciens (Kirby et al., 1997, 1998) and the XynC xylanase from F. succinogenes S85 each possess two catalytically active CDs. In contrast to the enzymes with a more complex structure, those produced by R. albus (Kirby et al., 1997) and the polycentric Orpinomyces usually have single CDs. The glycanase genes of the anaerobic ruminal fungi exhibit sequence conservation with their respective counterparts among the rumen bacteria, except for the celA gene of N. patriciarum as stated above (Chen et al., 1998; Denman et al., 1996). The similarity has led to the hypothesis that many of the fungal genes were acquired via horizontal gene transfer from the ruminal bacteria (Gilbert et al., 1992; Chen et al., 1997). Similarly, the protozoan cellulases and xylanases resemble those of the bacteria and fungi, for example, the P. multivesiculatum xylanase exhibited higher homology with family 11 xylanases from Gram-positive bacteria than with enzymes of eucaryotic origin (Devillard et al., 1999). This raises the possibility that the protozoan gene might have been acquired by horizontal gene transfer, perhaps from ingested bacteria. Superimposed upon gene transfer, evidence has been presented suggesting that genes encoding three mannanases of a Piromyces spp. (Millward-Sadler et al., 1996), two cellobiohydrolases of Orpinomyces PC-2 (Li et al., 1997b) and two glucanases from Orpinomyces (Chen et al., 1998) have originated via a gene duplication process. Gene duplication has also been observed in F. succinogenes as documented by the endoglucanases CelD and CelE (Malburg et al., 1996) and three new family 10 xylanase genes from F. succinogenes that are in series on the chromosome (J. Ha and C.W. Forsberg, unpublished data). It was hypothesized by Fanutti et al. (1995) that fungal mannanases and endoglucanases could have a common progenitor sequence. This is an interesting suggestion and perhaps is validated by endoglucanases and xylanases exhibiting both activities to varying degrees (Gilkes et al., 1991), and a xylanase exhibiting arabinosedebranching activity and a xylanase exhibiting arabinose-debranching activity (Matte and Forsberg, 1992).

88

C.W. Forsberg et al.

Structure and action of CBDs Few genes coding for glycosyl hydrolases with CBDs have been cloned from ruminal organisms. Those containing CBD sequences include the celA from N. patriciarum, celF from F. succinogenes S85 (Malburg et al., 1997), end1 from B. fibrisolvens H17C (Berger et al., 1989), and celA from Clostridium longisporum (Mittendorf and Thompson, 1993). One glucanase with a CBD has been cloned from each of E. caudatum and P. multivesiculatum as well. CBDs have one or more roles to play in the mechanism of cellulose hydrolysis. The CBD concentrates the attached enzyme at the cellulose surface. This was demonstrated by deletion of the CBD from CelA which reduced the Avicel cellulose hydrolysing activity by ninefold (Denman et al., 1996). In a separate study, attaching a CBD from Clostridium stercorarium to the R. albus endoglucanase IV was demonstrated to enhance cellulolysis at low cellulose concentrations, but not at high concentrations where enzyme molecules lacking the CBD had ready access to cleavage sites on the cellulose (Karita et al., 1996). The endoglucanase A (CenA) CBD from the non-rumen bacterium Cellulomonas fimi reportedly has the ability to disrupt noncovalent interactions between adjacent glucan chains in the cellulose microfibril (Din et al., 1991). A similar property has been reported for a scaffolding protein from C. cellulolyticum (Pagès et al., 1997). Thus some CBDs have the capacity to convert crystalline cellulose to amorphous strands at the surface. This may explain the ability of the Thermomonospora fusca E2 family II CBD to enhance the activity of the attached P. ruminicola glucanase against insoluble cellulose (Maglione et al., 1992) whereas a hybrid enzyme constructed from a Pseudomonas CBD and a Ruminococcus endoglucanase did not show enhanced catalytic activity (Poole et al., 1990). The cellulosome structure circumvents the need for individual glycosyl hydrolases to have CBDs. CBDs from different organisms are grouped into distinctive families on the basis of amino acid sequence similarity (Tomme et al., 1995). There are at least ten different families, with Type I and II being the most common domains in aerobic fungi and bacteria, respectively. A feature of many of these domains is a planar hydrophobic region in the surface, in which ligand binding is mediated by hydrophobic stacking interactions between solvent-exposed aromatic residues and the glucose-pyranose rings. There is a strong preference for insoluble forms of cellulose suggesting that the binding surface interacts with multiple cellulose chains. Binding by CBPs may in some cases seem practically irreversible. However, Jervis et al. (1997) tested the rate of surface diffusion of the C. fimi cellulases, Cex and CenA, and their respective CBDs using fluorescence recovery after photobleaching analysis. They reported that greater than 70% of bound molecules were mobile on the cellulose surface. A comparison with kinetics of crystalline cellulose hydrolysis suggested that surface diffusion rates did not limit cellulase activity. One can therefore visualize a model for the hydrolysis of plant cell wall by a cellulase with a CBD in which the CBD tethers the CD via a linker making available multiple cleavage sites. As the CBD diffuses, fresh sites become available along the path of diffusion, which presumably follow the orientation of the cellulose strands. No CBD from a ruminal organism has been fully characterized to date, but we can anticipate behaviour similar to their non-ruminal counterparts.

Microbial Adherence to the Plant Cell Wall

89

Regulation of hydrolase synthesis and activity It is commonly observed that glycosyl hydrolase synthesis by ruminal bacteria and fungi is enhanced by growth on cellulose and hemicellulose, and synthesis is reduced when cells are grown on glucose. This is well illustrated in a recent study by Bera et al. (1998) where they observed that seven cellulase genes, a cellodextrinase gene, a xylanase gene and a lichenase gene were over-expressed when F. succinogenes was grown on cellulose as compared with sugar-grown cultures. The expression of glycosyl-hydrolase genes of F. succinogenes therefore appears to be tightly regulated. The regulation of glycosyl hydrolases is less tightly controlled in P. ruminicola (Gardner et al., 1995). The xylanase was induced by xylan, but not by other sugars while the ␤-1,4-endoglucanases and mannanases were induced by xylan, but even higher activities were observed with mannose, cellobiose, or xylose. The endoglucanases and mannanases were repressed by sucrose, arabinose or rhamnose. Gardner et al. (1997) provided evidence that two endoglucanases and a mannanase were present in an operon and perhaps under the same control. Cotta et al. (1994) demonstrated the presence of cAMP in P. ruminicola, but it was practically absent from other anaerobes tested. They have since cloned a novel adenylate cyclase from P. ruminicola D31d (Cotta et al., 1998). However, the concentration of cAMP did not alter in organisms under conditions where catabolite repression-like phenomena occurred; therefore it may not have a role in regulation of hydrolase synthesis. Dalrymple and Swadling (1997) reported the presence of the cinR negative regulator of the cinB gene which codes for cinamoyl ester hydrolase in B. fibrisolvens. The addition of FAXX (O-[5-O-(trans-feruloyl)-␣-L-arabinofuranosyl]-(1,3)-O-␤-Dxylopyranosyl-(1,4)-D-xylopyranose) and Fara [5-O-(trans-feruloyl)-arabinofuranose], but not xylobiose or ferulic acid inhibited binding of CinR to DNA. This is the first characterized regulatory system in ruminal bacteria. A novel enhancer of cellulase action is produced by some ruminal organisms. O. joyonii produces a yellow affinity substance (YAS) that increases adsorption of cellulases to microcrystalline cellulose (Hodrová et al., 1998). A similar YAS is reportedly produced by R. flavefaciens (Kopecny and Hodrová, 1997). These substances have not been isolated, but a similar substance from C. thermocellum has a mass of 1036 Da and is comprised of a ring structure with a long hydrocarbon tail that may contain carbonyl and amino moieties (Ljungdahl et al., 1988). Although not previously recognized as a limitation on microbial digestion in the rumen, wheat plants produce a novel 20 kDa xylanase inhibitor that binds close to the active site of xylanase enzymes (McLauchlan et al., 1999). Whether other plants produce related inhibitor molecules remains to be determined.

Strategies to overcome limitations to the microbial degradation of plant cell walls From studies of plant cell wall structure it has become evident that a number of organizational features determine the nature of the plant cell wall biodegradation process. The most important of these features is the pore size distribution. This is essentially a

90

C.W. Forsberg et al.

measure of the spacing between the individual polymers contributing to the wall structure and is remarkably similar in all crop vegetation used for feed purposes. Direct measurement by a variety of probe-based methods has shown that most pores have a diameter between 2 and 4 nm (Chesson et al., 1997; Gardner et al., 1999). These dimensions are not sufficient to allow free diffusion into the wall by simple globular enzymes with masses greater than ~20 kDa. The porosity of the wall (Gardner et al., 1999) and its composition (Chesson et al., 1986) changes little during the course of degradation even when ~70% of dry matter has been eroded. The data are consistent with a pattern of degradation based on surface erosion imposed on the rumen bacteria by the impervious mixed polymer nature of the cell wall. Recent work showed that enzyme (cellulase and xylanase) supplementation of the diet can increase ruminal digestibility and milk production in cattle (Yang et al., 1999). This result is surprising considering the extensive array of potent endogenous fibrolytic enzymes produced by the rumen microflora. The most likely explanation for the beneficial effect is that addition of enriched extracellular polysaccharidases results in an immediate attack on freshly ingested plant material thereby providing additional available carbohydrate that encourages more rapid microbial growth, shortening the lag time required for microbial colonization. The net effect may be equivalent to a longer retention time within the rumen. However, one cannot preclude the possibility that the exogenous cellulases have more efficacious binding and catalytic properties. Evolution has not produced a solution that overcomes this limitation. Instead, most of the rumen bacteria and fungi have optimized a system of cell-wall degrading activities in the form of the cellulosome which recognizes the diffusion limitations imposed by the plant cell wall and which is highly adapted to a superficial mode of action. This makes it unlikely that the introduction of genes coding for single activities, as has been done in the past (see Forsberg et al., 1997), will contribute significantly to the degradation process. There remains the option of engineering the cellulosome itself (Bayer et al., 1998). However, more needs to be known about its structure and, in particular, the structure–function relationships of its component parts in the organism of choice, before this is likely to be effective. One possible option is the introduction of CBDs that have a greater avidity for cellulose, which disrupt the crystalline structure of the surface cellulose as shown by Din et al. (1991). Another is the introduction into the cellulosome structure of CDs with higher catalytic efficiency and, depending upon the organism of choice, ensuring that essential hemicellulases, for example, feruloly and coumaroyl esterase, arabinofuranosidase and acetylxylan esterase, are included as illustrated in Fig. 5.2. There may be more efficacious CBDs and CDs in non-ruminal organisms or those organisms, for example the ruminal fungi, with highly efficient cellulases, but which usually are a minor component of the population. With surface erosion as the predominant mechanism of bacterial (and rumen fungal?) cell wall breakdown, two factors are particularly important. Firstly, the amount of surface area available for colonization and, secondly, the chemistry of the available surface. Surface is created by feed processing and then by mastication/rumination, each opening plant cells and exposing the inner surface to colonization. Subsequent removal of polysaccharide from that surface can, in more lignified cell types, lead to the development of an inert surface in which any remaining polysaccharide is ‘protected’ from attack by the presence of phenolic compounds. Consequently, available surface area reaches a maximum and then diminishes with time resulting in the slowing of the rate

Microbial Adherence to the Plant Cell Wall

EG

91

CBD EGII

XynB

CBH

EGII

CBH

EGI

3

4

5

6

CBD 1

2

Man

XynA AXE

Ar

Bacterial or fungal cell

Man

CBD

XynB

7

CBH

EGII

CBH

EGI

CBD

9

8

XynA AXE

Ar

CBD

1

2

3

4

5

6

Man

XynB

CBH

EGII

CBH

EGI

1

2

3

4

5

6

CBD

7

8

XynA AXE

9

Ar

CBD

Cohesion domain (type I)

7

8

CBD

9

Cellulose-binding domain (CBD)

Dockerin domain (type I)

Catalytic domain

Dockerin domain (type II)

Linker

Fig. 5.2. Schematic representation of a cellulosome with high catalytic activity on plant cell walls. Symbols: Man, Mannanase; Xyn, xylanase; CBH, cellobiohydrolase; EG, endoglucanase; AXE, acetyl xylan esterase; Ar, arabinofuranosidase.

of degradation. The amount of surface that is denied to attacking organisms by this mechanism is a product of the amount of lignin present and the degree of its cross-linking to other wall polymers and defines the extent of degradation. The development of ruminal organisms able to digest lignin does not appear to be an option at present because of the aerobic requirements for the process to proceed at a rate compatible with the rate of passage. A strategy more likely to succeed is the modification of lignification in plants used as ruminant feeds. Attempts to achieve this by various pre-treatment methods, although partially successful, have not proven cost-effective.

92

C.W. Forsberg et al.

More permanent solutions through selective breeding and the use of recombinant DNA technology are under active investigation. Enzymes involved in the biosynthesis of lignin and tannin precursors have been targeted for genetic modification, and crop plants have been produced in which one or more of these enzymes have been down-regulated. One general observation made is that phenolic polymers appear much more plastic than was originally thought (Ralph, 1997). Modifying the nature of the precursors was found rarely to decrease the total amount of lignin formed but did have significant effects on its composition and properties (Boudet, 1998). Blocking a biochemical pathway at one point can also lead to a redirection of the flux of precursor molecules. Down-regulation of cinnamyl CoA reductase, an enzyme which catalyses the reduction of phenolic acids to the corresponding aldehyde, led to greater amounts of free ferulic acid present in the cell and increased significantly the number of diferulate cross-links within the wall (Piquemal et al., 1998). Because of the plasticity of lignin, targeting the extent of cross-linking between polymers within the wall may be a more effective way of altering cell wall degradation characteristics (Grabber et al., 1998). There is also an increasing appreciation of the importance to nutritional value of the spatial distribution of cell wall material within the plant. Unfortunately, the ability to apply this new understanding to the breeding of new crop varieties was, until recently, hampered by the lack of quantitative tools for the routine measurement of anatomy. Analysis of images from microscope sections now has revolutionized approaches for the measurement of anatomical features enabling the automated recognition of cell types and tissues to be used as selection criteria (Travis et al., 1996). However, the lack of knowledge of the genetic basis to anatomy has meant that development of improved crop plants using anatomical features as selection criteria is restricted to conventional breeding at present.

Concluding remarks As we look to the future there would appear to be no short-term opportunities to develop genetically modified ruminal organisms that will radically improve the rate and extent of plant cell wall digestion. However, furthering our understanding of the structure of the cellulosome and related cellulase systems of ruminal organisms and comparative studies on the CBDs and CDs of ruminal and other organisms may reveal new opportunities to improve the catalytic properties of the ruminal cellulases. This strategy, in conjunction with continued exploration of genetic methods to reduce the lignin content of fibrous plants used as ruminant feeds, should eventually give rise to improved digestion of plant components that at present are considered to be largely recalcitrant to ruminal digestion.

Acknowledgements The authors thank E.A. Bayer, P. Beguin, J.-P. Belaich, J. Ha, P. Mosoni and J.H.D. Wu for communicating data prior to publication, and L.B. Selinger for providing journal articles not available at the authors’ institutions.

Microbial Adherence to the Plant Cell Wall

93

References Ali, B.R.S., Zhou, L.Q., Graves, F.M., Freedman, R.B., Black, G.W., Gilbert, H.J. and Hazlewood, G.P. (1995) Cellulases and hemicellulases of the anaerobic fungus Piromyces constitute a multiprotein cellulose-binding complex and are encoded by multigene families. FEMS Microbiology Letters 125, 15–21. Bayer, E.A., Chanzy, H., Lamed, R. and Shoham, Y. (1998) Cellulose, cellulases and cellulosomes. Current Opinion in Structural Biology 8, 548–557. Beguin, P. and Lemaire, M. (1996) The cellulosome: an exocellular, multiprotein complex specialized in cellulose degradation. Critical Review of Biochemistry and Molecular Biology 31, 201–236. Bera, C., Gaudet, G. and Forano, E. (1998) Regulation of glycosyl-hydrolase genes expression in Fibrobacter succinogenes S85. In: Ohmiya, K., Hayashi, K., Sakka, K., Kobayashi, Y., Karita, S. and Kimura, T. (eds) Genetics, Biochemistry and Ecology of Cellulose Degradation. Uni Publishers Co., Ltd, Japan, pp. 541–544. Berger, E., Jones, W.A., Jones, D.T. and Woods, D.R. (1989) Cloning and sequencing of an endoglucanase (end1) gene from Butyrivibrio fibrisolvens H17c. Molecular and General Genetics 219, 193–198. Blair, B.J. and Anderson, K. (1998) Comparison of staining techniques for scanning electron microscopic detection of ultrastructural protuberances on cellulolytic bacteria. Biotechnic and Histochemistry 73, 107–113. Boudet, A.-M. (1998) A new view of lignification. Trends in Plant Science 3, 67–71. Chen, H., Li, X.-L. and Ljungdahl, L.G. (1997) Sequencing of a 1,3–1,4-␤-D-glucanase (lichenase) from the anaerobic fungus Orpinomyces strain PC-2: Properties of the enzyme expressed in Escherichia coli and evidence that the gene has a bacterial origin. Journal of Bacteriology 179, 6028–6034. Chen, H., Li, X.-L., Blum, D.L. and Ljungdahl, L.G. (1998) Two genes of the anaerobic fungus Orpinomyces sp. strain PC2 encoding cellulases with endoglucanase activities may have arisen by gene duplication. FEMS Microbiology Letters 159, 63–68. Chesson, A. and Forsberg, C.W. (1997) Polysaccharide degradation by rumen microorganisms. In: Hobson, P. and Stewart, C. (eds) The Rumen Microbial Ecosystem, 2nd edn. Chapman & Hall Ltd, Andover, UK, pp. 329–381. Chesson, A., Stewart, C.S., Dalgarno, K. and King, T.P. (1986) Degradation of isolated grass mesophyll, epidermis and fibre cell walls in the rumen and by cellulolytic rumen bacteria in axenic culture. Journal of Applied Bacteriology 60, 327–336. Chesson, A., Gardner, P.T. and Wood, T.J. (1997) Cell wall porosity and available surface area of wheat straw and wheat grain fractions. Journal of the Science of Food and Agriculture 75, 289–295. Cotta, M.A., Wheeler, M.B. and Whitehead, T.R. (1994) Cyclic AMP in ruminal and other anaerobic bacteria. FEMS Microbiology Letters 124, 355–359. Cotta, M.A., Whitehead, T.R. and Wheeler, M.B. (1998) Identification of a novel adenylate cyclase in the ruminal anaerobe, Prevotella ruminicola D31d. FEMS Microbiology Letters 164, 257–260. Dalrymple, B.P. and Swadling, Y. (1997) Expression of a Butyrivibrio fibrisolvens E14 gene (cinB) encoding an enzyme with cinnamoyl ester hydrolase activity is negatively regulated by the product of an adjacent gene (cinR). Microbiology 143, 1203–1210. Dehority, B.A. and Tirabasso, P.A. (1998) Effect of ruminal cellulolytic bacterial concentrations on in situ digestion of forage cellulose. Journal of Animal Science 76, 2905–2911. Denman, S., Xue, G.P. and Patel, B. (1996) Characterization of a Neocallimastix patriciarum cellulase cDNA (celA) homologous to Trichoderma reesei cellobiohydrolase II. Applied and Environmental Microbiology 62, 1889–1896. Devillard, E., Newbold, C.J., Scott, K.P., Forano, E., Wallace, R.J., Jouany, J.-P. and Flint, H.J.

94

C.W. Forsberg et al. (1999) A xylanase produced by the rumen anaerobic protozoan Polyplastron multivesiculatum shows close sequence similarity with family 11 xylanases from Gram-positive bacteria. FEMS Microbiology Letters 181, 145–152. Dijkerman, R., Opdencamp, H.J.M., Vanderdrift, C. and Vogels, G.D. (1997) The role of the cellulolytic high molecular mass (HMM) complex of the anaerobic fungus Piromyces sp. strain E2 in the hydrolysis of microcrystalline cellulose. Archives of Microbiology 167, 137–142. Dijkstra, J. and Tamminga, S. (1995) Simulation of the effects of diet on the contribution of rumen protozoa to degradation of fibre in the rumen. British Journal of Nutrition 74, 617–634. Din, N., Gilkes, N.R., Tekant, B., Miller, R.C., Jr, Warren, R.A.J. and Kilburn, D.G. (1991) Non-hydrolytic disruption of cellulose fibres by the binding domain of a bacterial cellulase. Bio/Technology 9, 1096–1099. Ding, S.-Y., Bayer, E.A., Shoham, Y. and Lamed, R. (1999) Preliminary evidence of high-molecular-weight scaffodin-like proteins from Ruminococcus flavefaciens. P5.12, 3rd Carbohydrate Bioengineering Meeting, University of Newcastle Upon Tyne 11–14 April. Doerner, K.C. and White, B.A. (1990) Assessment of the endo-ß-1,4-glucanase components of Ruminococcus flavefaciens FD1. Applied and Environmental Microbiology 56, 1844–1850. Fanutti, C., Ponyi, T., Black, G.W., Hazlewood, G.P. and Gilbert, H.J. (1995) The conserved noncatalytic 40-residue sequence in cellulases and hemicellulases from anaerobic fungi functions as a protein docking domain. Journal of Biological Chemistry 270, 29314–29322. Forsberg, C.W., Cheng, K.-J. and White, B.A. (1997) Polysaccharide degradation in the rumen and large intestine. In: Mackie, R.I. and White, B.A. (eds) Gastrointestinal Microbiology. Chapman & Hall, New York, pp. 319–379. Gardner, R.G., Wells, J.E., Russell, J.B. and Wilson, D.B. (1995) The effect of carbohydrates on the expression of the Prevotella ruminicola ␤-1,4-D-endoglucanase. FEMS Microbiology Letters 125, 305–310. Gardner, R.G., Wells, J.E., Fields, M.W., Wilson, D.B. and Russell, J.B. (1997) A Prevotella ruminicola B14 operon encoding extracellular polysaccharide hydrolases. Current Microbiology 35, 274–277. Gardner, P.T., Wood, T.J., Chesson, A. and Stuchbury, T. (1999) Effect of degradation on the porosity and surface area of forage cell walls of differing lignin content. Journal of the Science of Food and Agriculture 79, 11–18. Gilbert, H.J., Hazlewood, G.P., Laurie, J.I., Orpin, C.G. and Xue, G.P. (1992) Homologous catalytic domains in a rumen fungal xylanase: evidence for gene duplication and prokaryotic origin. Molecular Microbiology 6, 2065–2072. Gilkes, N.R., Henrissat, B., Kilburn, D.G., Miller, J.R.C. and Warren, R.A.J. (1991) Domains in microbial ␤-1,4-glycanases: sequence conservation, function, and enzyme families. Microbiological Reviews 55, 303–315. Gong, J. and Forsberg, C.W. (1989) Factors affecting adhesion of Fibrobacter succinogenes subsp. succinogenes and adherence-defective mutants to cellulose. Applied and Environmental Microbiology 55, 3039–3044. Gong, J., Egbosimba, E.E. and Forsberg, C.W. (1996) Cellulose binding proteins of Fibrobacter succinogenes and the possible role of a 180-kDa cellulose-binding glycoprotein in adhesion to cellulose. Canadian Journal of Microbiology 42, 453–460. Grabber, J.H., Hatfield, R.D. and Ralph, J. (1998) Diferulate cross-links impede the enzymatic degradation of non-lignified maize walls. Journal of the Science of Food and Agriculture 77, 193–200. Groleau, D. and Forsberg, C.W. (1983) Solubilization of the extracellular membraneous carboxymethylcellulase activity produced by the rumen bacterium Bacteroides succinogenes. Canadian Journal of Microbiology 29, 710–717. Henrissat, B. (1998) Enzymology of cell-wall degradation. Biochemical Society Transactions 26, 153–156.

Microbial Adherence to the Plant Cell Wall

95

Henrissat, B., Teere, T.T. and Warren, R.A.J. (1998) A scheme for designating enzymes that hydrolyse the polysaccharides in the cell walls of plants. FEBS Letters 425, 352–354. Hodrová, B., Kopecny, J. and Kas, J. (1998) Cellulolytic enzymes of rumen anaerobic fungi Orpinomyces joyonii and Caecomyces communis. Research in Microbiology 149, 417–427. Jervis, E., Haynes, C. and Kilburn, D. (1997) Surface diffusion of cellulases and their isolated binding domains on cellulose. Journal of Biological Chemistry 272, 24016–24023. Karita, S., Sakka, K. and Ohmiya, K. (1996) CBDs confer an enhanced activity against insoluble cellulose to Ruminococcus albus endoglucanase IV. Journal of Fermentation Biotechnology 81, 553–556. Karita, S., Sakka, K. and Ohmiya, K. (1997) Cellulosomes, cellulase complexes of anaerobic microbes: their structure models and functions. In: Onodera, R., Itabashi, H., Ushida, K., Yano, H. and Sasuki, Y. (eds), Rumen Microbes and Digestive Physiology in Ruminants. Japan Scientific Societies Press, Tokyo/S. Karger, Basel, pp. 47–57. Kirby, J., Martin, J.C., Daniel, A.S. and Flint, H.J. (1997) Dockerin-like sequences in cellulases and xylanases from the rumen cellulolytic bacterium Ruminococcus flavefaciens. FEMS Microbiology Letters 149, 213–219. Kirby, J., Aurilia, V., McCrae, S.I., Martin, J.C. and Flint, H.J. (1998) Plant cell wall degrading enzyme complexes from the cellulolytic rumen bacterium Ruminococcus flavefaciens. Biochemical Society Transactions 26, S169. Kopecny, J. and Hodrová, B. (1997) The effect of yellow affinity substance on cellulases of Ruminococcus flavefaciens. Letters in Applied Microbiology 25, 191–196. Krause, D.O., Dalrymple, B.P., Smith, W.J., Mackie, R.I. and McSweeney, C.S. (1999) 16S rDNA sequencing of Ruminococcus albus and Ruminococcus flavefaciens: design of a signature probe and its application in adult sheep. Microbiology 145, 1797–1807. Lamed, R., Naimark, J., Morgenstern, E. and Bayer, E.A. (1987) Specialized cell surfaces structures in cellulolytic bacteria. Journal of Bacteriology 169, 3792–3800. Leatherwood, J.M. (1973) Cellulose degradation by Ruminococcus. Federation Proceedings 32, 1815–1819. Li, X.-L., Chen, H. and Ljungdahl, L.G. (1997a) Monocentric and polycentric anaerobic fungi produce structurally related cellulases and xylanases. Applied and Environmental Microbiology 63, 628–635. Li, X.-L., Chen, H. and Ljungdahl, L.G. (1997b) Two cellulases, CelA and CelC, from the polycentric anaerobic fungus Orpinomyces strain PC2 contain N-terminal docking domains for a cellulase–hemicellulase complex. Applied and Environmental Microbiology 63, 4721–4728. Lin, L.L. and Thomson, J.A. (1991) An analysis of the extracellular xylanases and cellulases of Butyrivibrio fibrisolvens H17c. FEMS Microbiology Letters 84, 197–204. Ljungdahl, L.G., Coughlan, M.P., Mayer, F., Mori, Y., Hon-Nami, H. and Hon-Nami, K. (1988) Macrocellulase complexes and yellow affinity substance from Clostridium thermocellum. Methods in Enzymology 160, 483–500. McIntosh, L.P., Hand, G., Johnson, P.E., Joshi, M.D., Körner, M., Plesniak, L.A., Ziser, L., Wakarchuk, W.W. and Withers, S.G. (1996) The pKa of the general acid/base carboxyl group of a glycosidase cycles during catalysis: A 13C-NMR study of Bacillus circulans xylanase. Biochemistry 35, 9958–9966. McLauchlan, W.R., Garcia-Conesa, M.T., Williamson, G., Rosa, M., Ravestein, P. and Maat, J. (1999) A novel class of protein from wheat which inhibits xylanases 1. Biochemical Journal 338, 441–446. Maglione, G., Matsushita, O., Russell, J.B. and Wilson, D.B. (1992) Properties of a genetically reconstructed Prevotella ruminicola endoglucanase. Applied and Environmental Microbiology 58, 3593–3597. Malburg, L.M., Iyo, A.H. and Forsberg, C.W. (1996) A novel family 9 endoglucanase gene (celD), whose product cleaves substrates mainly to glucose, and its adjacent upstream homolog (celE) from Fibrobacter succinogenes S85. Applied and Environmental Microbiology 62, 898–906.

96

C.W. Forsberg et al. Malburg, S.R.C., Malburg, L.M., Liu, T., Iyo, A.H. and Forsberg, C.W. (1997) Catalytic properties of the cellulose binding endoglucanase F from Fibrobacter succinogenes S85. Applied and Environmental Microbiology 63, 2449–2453. Matte, A. and Forsberg, C.W. (1992) Purification, characterization, and mode of action of endoxylanases 1 and 2 from Fibrobacter succinogenes S85. Applied and Environmental Microbiology 58, 157–168. Millward-Sadler, S.J., Hall, J., Black, G.W., Hazlewood, G.P. and Gilbert, H.J. (1996) Evidence that the Piromyces gene family encoding end-1,4-mannanases arose through gene duplication. FEMS Microbiology Letters 141, 183–188. Miron, J. and Forsberg, C.W. (1998) Features of Fibrobacter intestinalis DR7 mutant which is impaired with its ability to adhere to cellulose. Anaerobe 7, 35–48. Miron, J. and Forsberg, C.W. (1999) Characterisation of cellulose binding proteins which are involved in adhesion mechanism of Fibrobacter intestinalis DR7. Applied Microbiology and Biotechnology 51, 491–497. Miron, J., Yokoyama, M.T. and Lamed, R. (1989) Bacterial cell surface structures involved in lucerne cell wall degradation by pure cultures of cellulolytic rumen bacteria. Applied Microbiology and Biotechnology 32, 218–222. Miron, J., Morag, E., Bayer, E.A., Lamed, R. and Ben-Ghedalia, D. (1998) An adhesion defective mutant of Ruminococcus albus SY3 is impaired in its capability to degrade cellulose. Journal of Applied Microbiology 84, 249–254. Mitsumori, M. and Minato, H. (1995) Distribution of cellulose-binding proteins among the representative strains of rumen bacteria. Journal of General and Applied Microbiology 41, 297–306. Mitsumori, M., Minato, H., Sekizaki, T., Uchida, I. and Ito, H. (1996) Cloning, nucleotide sequence and expression of the gene encoding the cellulose-binding protein 1 (CBP1) of Fibrobacter succinogenes S85. FEMS Microbiology Letters 139, 43–50. Mittendorf, V. and Thomson, J.A. (1993) Cloning of an endo-(1–4)-␤-glucanase gene, celA, from the rumen bacterium Clostridium sp. (C. longisporum) and characterization of its product, CelA, in Escherichia coli. Journal of General Microbiology 139, 3233–3242. Morris, E.J. and Cole, O.J. (1987) Relationship between cellulolytic activity and adhesion to cellulose in Ruminococcus albus. Journal of General Microbiology 133, 1023–1032. Mosoni, P. (1999) Characterization of two spontaneous adherence defective mutants of Ruminococcus albus strains. South African Journal of Animal Science 29, 78–79. Ohara, H., Karita, S., Kimura, T., Sakka, K. and Ohmiya, K. (1998) Multicellulase complex of Ruminococcus albus F40. Proceedings of the MIE Bioforum 1998 Genetics, Biochemistry and Ecology of Cellulose Degradation, Suzuka, Japan, p. 140. Orpin, C.G. (1983) The role of ciliate protozoa and fungi in the rumen digestion of plant cell walls. Animal Feed Science and Technology 10, 121–143. Pagès, S., Gal, L., Bélaïch, A., Gaudin, C., Tardif, C. and Bélaïch, J.-P. (1997) Role of scaffolding protein CipC of Clostridium cellulolyticum in cellulose degradation. Journal of Bacteriology 179, 2810–2816. Pegden, R.S., Larson, M.A., Grant, R.J. and Morrison, M. (1998) Adherence of the gram positive bacterium Ruminococcus albus to cellulose and identification of a novel form of cellulose-binding protein which belong to the Pil family of proteins. Applied and Environmental Microbiology 180, 5921–5927. Piquemal, J., Lapierre, C., Myton, K., O’Connell, A., Schuch, W., Grima-Pettenati, J. and Boudet, A.-M. (1998) Down-regulation of cinnamoyl-CoA reductase induces significant changes of lignin profiles in transgenic tobacco plants. Plant Journal 13, 71–83. Poole, D.H., Hazlewood, G.P., Laurie, J.I., Barker, P.J. and Gilbert, H.J. (1990) Nucleotide sequence of the Ruminococcus albus SY3 endoglucanase genes celA and celB. Molecular and General Genetics 223, 217–223.

Microbial Adherence to the Plant Cell Wall

97

Ralph, J. (1997) Fake plastic trees. Chemistry and Industry, p. 708. Stack, R.J. and Hungate, R.E. (1984) Effect of 3-phenylpropionic acid on capsule and cellulases of Ruminococcus albus 8. Applied and Environmental Microbiology 48, 218–223. Takenaka, A., D’Silva, C.G., Kudo, H., Itabashi, H. and Cheng, K.-J. (1999) Molecular cloning, expression, and characterization of an endo-␤-1,4-glucanase cDNA from Epidinium caudatum. Journal of General and Applied Microbiology 45, 57–61. Tomme, P., Warren, R.A.J. and Gilkes, N.R. (1995) Cellulose hydrolysis by bacteria and fungi. Advances in Microbial Physiology 37, 1–81. Travis, A.J., Hirst, D.J. and Chesson, A. (1996) Automatic classification of plant cells according to tissue type using anatomical features obtained by the distance transform. Annals of Botany 78, 325–331. Trinci, A.P.J., Davies, D.R., Gull, K., Lawrence, M.I., Nielsen, B.B., Rickers, A. and Theodorou, M.K. (1994) Anaerobic fungi in herbivorous animals. Mycological Research 98, 129–152. Weimer, P.J., Waghorn, G.C., Odt, C.L. and Mertens, D.R. (1999) Effect of diet on populations of three species of ruminal cellulolytic bacteria in lactating dairy cows. Journal of Dairy Science 82, 122–134. Wilson, J.R. and Mertens, D.R. (1995) Cell wall accessibility and cell structure limitations to microbial digestion of forage. Crop Science 35, 251–259. Wood, T.M., Wilson, C.A. and Stewart, C.S. (1982) Preparation of the cellulase from the cellulolytic anaerobic rumen bacterium Ruminococcus albus and its release from the bacterial cell wall. Biochemical Journal 205, 129–137. Wood, T.M., Wilson, C.A. and McCrae, S.I. (1995) The cellulase system of the anaerobic rumen fungus Neocallimastix frontalis: studies on the properties of fractions rich in endo (1,4)-ß-Dglucanase activity. Applied Microbiology and Biotechnology 44, 177–184. Xue, G.P., Orpin, C.G., Gobius, K.S. and Orpin, C.G. (1992) A novel polysaccharide hydrolase cDNA (celD) from Neocallimastix patriciarum encoding three multi-functional catalytic domains with high endoglucanase, cellobiohydrolase and xylanase activities. Journal of General Microbiology 138, 2397–2403. Yang, W.Z., Beauchemin, K.A. and Rode, L.M. (1999) Effects of an enzyme feed additive on extent of digestion and milk production of lactating dairy cows. Journal of Dairy Science 82, 391–403.

6

The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism M. MORRISON Department of Animal Science, School of Biological Sciences and Center for Biotechnology, University of Nebraska, Lincoln, Nebraska, USA

Introduction The relationships between the ruminant animal and its resident microflora, and the ensuing impacts on protein nutrition, have been examined in detail for more than 40 years. The foundational studies such as those by Hogan and co-workers demonstrated the interface between host and microbial metabolism of non-protein nitrogen sources; and the practical ramifications of this interface were clearly demonstrated in the widely referenced study by Virtanen (1966). Egan (1965) helped establish another precedent in our understanding of ruminant protein nutrition. By showing that positive responses in animal physiology coincided with the intra-abomasal infusion of casein, their studies serve as the underpinning of the ‘by-pass protein’ concept. To this day, much of the practical interest in ruminal nitrogen metabolism relates to the optimization of microbial growth with different nitrogen sources, as well as the optimization of postruminal supply of microbial and feed proteins. Additionally, as the relative amount and/or biological value of feed protein increases, there is a shift in research interest, away from the biosynthetic processes of ruminal microbes, and more towards their degradative processes. For these reasons, the ruminal microflora continues to provide a unique and varied set of challenges for nutritionists and microbiologists intent on improving the protein nutrition of animals, in both extensive and intensive production scenarios. Hungate (1960) considered that a meaningful analysis of any microbial habitat requires an understanding of: (i) the types of kinds of microorganisms present (ecology); (ii) the activities possessed by these microorganisms (enzymology); and (iii) factors affecting the expression of these activities (regulation). There is no shortage of recent reviews that provide a detailed accounting of our current understanding of these aspects, from an organismal and biochemical perspective (Morrison and Mackie, 1996; Cotta and Russell, 1997; Wallace et al., 1998 ). This review will attempt to take a different approach to the topic, by focusing on recent and pending advances contributed via genetics and molecular biology. In addition to providing new information © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

99

M. Morrison

100

relating to the ecology and enzymology of ruminal nitrogen metabolism, these techniques are likely to help elucidate how nitrogen metabolism is coordinated, and what other physiological process(es) ensue. Ultimately, a more precise identification of the rate-limiting parameter(s) of ruminal nitrogen metabolism, from an ecological, enzymatic, or regulatory perspective, should expedite advances in ruminant protein nutrition.

A molecular-based perspective of microbial ecology and ruminal nitrogen metabolism: cultivation-dependent and cultivation-independent studies The value from examination of ruminal microorganisms in pure or defined mixed cultures can not be overstated. For instance, the principle of interspecies hydrogen transfer, foundational to our understanding of the anaerobic metabolism of organic matter, arose from such studies. However, most microbiologists and nutritionists interested in the ruminal habitat have also wondered how representative laboratory isolates are of the microbial diversity inherent to the rumen. Additionally, how representative is the growth and metabolism of ruminal microorganisms under laboratory conditions, relative to their behaviour in their ‘real world’? Various molecular-based technologies are now available and are being gradually adapted for use with ruminal microorganisms. Although an inherent advantage of these technologies will be their utilization for in situ examination of ruminal microbiology and ecology, their greatest impact in ruminal microbiology is currently arising from their integration with other microbiological methods. Some examples with specific relevance to ruminal nitrogen metabolism are outlined below.

Proteolytic and obligate peptide- and amino-acid-fermenting bacteria Probably the most salient development in ruminal microbiology since the isolation and characterization of the anaerobic fungi has been the identification of obligate peptideand amino-acid-fermenting bacteria. Chen and Russell (1988, 1989) obtained three isolates of Gram-positive bacteria not only with very high specific activities of ammonia production, but able to grow rapidly with peptides or amino acids as their sole energy source. Although two of the isolates were ultimately identified by taxonomic characteristics as Peptostreptococcus anaerobius and Clostridium sticklandii, the third isolate possessed taxonomic characteristics which did not clearly match the criteria necessary for assignment to established bacterial species. This discrepancy was resolved by the use of molecular techniques; in this case, 16S rRNA samples were used as templates for primer elongation by reverse transcriptase, and nearly complete sequences were aligned with other sequences available in the databases. On the basis of these studies the third isolate warranted designation as a new species and, following the isolation of additional strains with similar phenotypic and phylogenetic characteristics, the strains were classified as Clostridium aminophilum sp. nov. (Paster et al., 1993). Although the ecological significance of these types of bacteria in ruminal nitrogen metabolism was initially questioned by some, bacteria with similar ecophysiological

The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism

101

properties have been isolated from ruminants in other parts of the world. Attwood et al. (1998) recently isolated from New Zealand ruminants a variety of bacteria with high specific rates of ammonia production. Five of these strains are phenotypically similar to the isolates identified by Chen and Russell (1988, 1989), at least in terms of sensitivity to monensin and their growth being restricted primarily to small peptides or amino acids. Phylogenetic (16S rRNA) analyses placed two of the isolates in the genus Peptostreptococcus, one in the genus Eubacterium, and two more in the family Bacteroideacea (presumably, the first Gram-negative bacteria of this type). Wallace et al. (1999) have also described the biochemical and phylogenetic characteristics of a Grampositive rod named Eubacterium pyruvovorans sp. nov., which is capable of growth with either peptides, amino acids, or organic acids such as pyruvate or oxaloacetate. In addition to acetate, this bacterium also produced long- and branched-chain organic acids such as valerate and caproate during growth. It is now widely accepted that bacteria with high specific rates of ammonia production are not uncommon in ruminant animals and, in some instances, might comprise a significant percentage of the culturable bacteria in the rumen. Despite the breakthroughs in our understanding of ruminal ammonium production arising from these cultivation-dependent studies, there are some inherent difficulties with such an approach. The culturable number of these bacteria can be relatively low (~107 per gram of ruminal contents; Chen and Russell, 1989) and the lack of completely selective media can make it difficult to accurately quantify their numbers (Attwood et al., 1998). To overcome these difficulties, molecular techniques have been developed and utilized effectively for population analysis, and with a variety of production/nutritional scenarios. Krause and Russell (1996) were the first to use a combination of in vivo sampling and a continuous culture system inoculated with predominant ruminal bacteria (PRB), to examine how monensin influences the persistence and abundance of the obligate amino-acid-fermenting bacteria. Despite all three strains being monensin sensitive in pure culture under batch conditions, C. aminophilum was found to persist and increase in relative abundance, both in vivo and within in vitro continuous cultures containing PRB, in the presence of monensin. Although other reasons have been forwarded to explain the ‘protein sparing’ effect of monensin, the conclusions reached by Krause and Russell (1996) also seem valid: monensin does not have a greater impact on ammonia production kinetics because of the persistence of C. aminophilum. It will be interesting to see if similar patterns of persistence exist for any of the other types of obligate amino-acid-fermenting bacteria isolated elsewhere. Studies such as these also highlight that while much can be learned from pure culture studies, microorganisms can behave quite differently when cultivated as part of a more complex consortia. Another molecular-based technology which circumvents cultivation-dependent analysis of ruminal microbiology is competitive PCR (cPCR). The technique has been used to enumerate the abundance of the proteolytic bacterium Clostridium proteoclasticum in New Zealand ruminants (Reilly and Attwood, 1998). Under laboratory conditions, this bacterium produces extremely high levels of proteinase activity, and 16S rRNA analysis showed it is very similar to several bacterial strains currently classified within the genus Butyrivibrio fibrisolvens. The cPCR technique is dependent on the development of a ‘competitor DNA’ template, which is usually generated by removal of a restriction enzyme fragment within a clone of a 16S rRNA gene isolated from the bacterium

M. Morrison

102

being studied. Primers specific only for the intact gene (‘target’ DNA) and the competitor template are then used in PCR reactions, the products are resolved by agarose gel electrophoresis and the log ratio of band intensities for the target and competitor PCR products are determined. A standard curve is first generated by combining serial dilutions of cells containing the target DNA with fixed amounts of competitor DNA, and the log ratio of PCR band intensities is plotted relative to cell numbers. DNA from environmental samples are then ‘spiked’ with a known quantity of the competitor template and subjected to PCR. In this manner, the quantity of target DNA in an environmental sample can be quantified, and in the case of C. proteoclasticum, as few as 2500 cells g⫺1 of ruminal contents can be detected. Reilly and Attwood (1998) showed the numbers of C. proteoclasticum and closely related species did not vary greatly among animals consuming rations with varying levels of carbohydrate and/or protein. These type of studies have now been expanded to include a greater range of proteolytic bacteria, including Streptococcus, Butyrivibrio and Eubacterium spp. (Attwood and Klieve, 2000). Interestingly, although the numbers of Streptococcus spp. were unchanged in response to diet, the numbers of Butyrivibrio were stimulated, and the numbers of Eubacterium spp. suppressed, by carbohydrate supplementation. Similar studies are now underway for populations of obligate amino-acid-fermenting bacteria (G.T. Attwood, personal communication), and it will be interesting to see whether the fluctuations in proteolytic bacteria in response to plane of nutrition extend also to this group of ruminal bacteria.

Examination of intraruminal nitrogen recycling Since Nolan and Leng (1972) showed that 30–50% of rumen microbial protein may recycle through the ammonium-N pool, ecological and physiological explanations for this intraruminal N-recycling have been sought. Attention in the past has been directed largely towards quantifying the impact of ruminal protozoa on this phenomenom, primarily via the use of defaunation agents. In general terms, elimination of ruminal protozoa corresponds with increased bacterial yield and improved protein outflow from the rumen, provided there is limited effect from defaunation on the ruminal fermentation of organic matter (Jouany et al., 1988). Additionally, Ffoulkes and Leng (1989) estimated that as much as 74% of the protozoal biomass never leaves the rumenreticulum. However, although this may be true for forage-fed animals, evidence suggests that ruminal protozoa not only slow down the rate of ruminal starch digestion in grain-fed animals, but shift the site of its digestion to the small intestine, as starchladen protozoa are removed from the rumen (Mendoza et al., 1993). Given the inherent difficulties associated with microscopic examinations to quantify and differentiate ruminal protozoa, as well as with the use of biochemical markers (such as diaminopimelic and aminoethylphosphonic acids) to quantify flow rates of protozoal and bacterial nitrogen from the rumen, molecular methods perhaps offer alternative approaches. The genes encoding small subunit rRNAs, and one or two other highly conserved structural proteins, have now been isolated from cDNA libraries of several predominant rumen protozoa (e.g. Wright et al., 1997; Eschenlauer et al., 1998). It should be possible to extend the application of these clones beyond phylogenetic analyses, and develop probes to quantify the relative abundance of particular groups of ruminal protozoa. One interesting example would be to examine the degree to which

The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism

103

protozoal lysis contributes to intraruminal N-recycling, and whether the elevated level of ammonium present in faunated animals is attributable more to protozoal predation of bacteria, or protozoal lysis as suggested by Wells and Russell (1996). More recently, the relationship(s) between intraruminal N-recycling and bacteriophage populations has received some attention, but unfortunately, efforts to understand this component of ruminal ecology are not widespread. Largely due to the efforts of Klieve and co-workers, molecular technologies have provided and may continue to provide relevant breakthroughs in our understanding of bacteriophage biology and population dynamics. Bacteriophages capable of infecting some of the better-known ruminal bacteria have been isolated and partially characterized (Klieve et al., 1989) and both lytic and temperate bacteriophages have been identified (Lockington et al., 1988; Kleive et al., 1989, 1991). All known ruminal bacteriophages possess a single copy of a double-stranded DNA molecule, which has facilitated the development of techniques to examine genetic diversity and bacteriophage abundance in response to diet and feeding behaviour. Klieve and Swain (1993) developed a method of harvesting bacteriophages from ruminal fluid coupled with the lysis of the bacteriophage capsid and separation of bacteriophage DNA by pulsed-field gel electrophoresis (PFGE). Genome size can be estimated by the migration distance of the DNA molecules in PFGE gels, relative to a set of DNA standards of known size. Additionally, changes in the abundance of DNA within a particular region of the gel can be quantified by laser densitometry, and the diurnal variation in bacteriophage numbers can be estimated from these changes in DNA staining intensity. Using these methods, Klieve and Swain (1993) found bacteriophage DNA could be subdivided into two main components. The bulk of the DNA, postulated to have been derived from temperate bacteriophages, was present as a broad band ranging from 30 to 200 kb. Distinct bands of DNA, both small (~10 kb) and very large (~850 kb) in size, were also apparent and postulated to represent blooms of lytic bacteriophage. Using these same methods, Swain et al. (1996) also showed there is a marked diurnal fluctuation in total bacteriophage numbers, the lowest concentrations occurring within two hours of feeding, and maximal concentrations occuring 8–10 h later. In addition to this daily fluctuation in numbers, Klieve et al. (1998) found that animals consuming green pasture tend to have the highest concentration of bacteriophages, and the concentrations were twofold and tenfold lower in animals consuming dry forages, and grain-fed animals, respectively. However, whether these differences are the result of sampling time relative to feeding, or other factors such as bacterial resistance to bacteriophage infection, are not clear. With the development of these techniques the next logical step appears to be to superimpose measurements of the intraruminal turnover of nitrogen among ammonium and non-ammonium pools upon quantitative measurements of fluctuations in protozoal, bacterial and bacteriophage populations. Accordingly, the relative contributions from each of these biotic components of the ruminal habitat to intraruminal nitrogen recycling, and the physicochemical factor(s) underpinning the phenomenon, should be elucidated. Such information is most likely necessary for the development of new, productive methods of curbing intraruminal nitrogen recycling. In summation, the molecular methods outlined above are providing the means necessary to examine rumen microbial diversity and phylogeny in more detail. The transition of these technologies from providing information of fundamental microbiological significance to the realm of providing information of a more pragmatic nature is

M. Morrison

104

gradually occurring. These technologies also provide the opportunity for detailed, quantitative examination of microorganisms in situ and their response to fluctuations in nutrient supply.

A genetics-based approach to the examination of microbial physiology and ruminal nitrogen metabolism New insights into ruminal microbiology, especially in relation to fibre degradation, have been forthcoming from the use of recombinant DNA technology (Flint, 1997). However, application of the technology has also been one-sided, limited largely to the characterization of clones stably maintained and expressed in Escherichia coli. Although heterologous gene expression in ruminal bacteria still remains limited and evasive, other genetics-based methods of analysis are being used to generate a clearer understanding of ruminal nitrogen metabolism. Two methods finding recent application in ruminal microbiology will be described here: mutational analysis and differential display reverse transcriptase PCR (DD RT-PCR).

Mutational analysis of nitrogen metabolism in Prevotella bryantii Mutagenesis procedures, despite being used extensively in microbiological studies, have found little application to-date in the field of ruminal microbiology. Newbold et al. (1992) described the properties of ionophore-resistant Prevotella (Bacteroides) ruminicola cells enriched by cultivation in the presence of increasing concentrations of tetronasin; and spontaneous mutants of P. ruminicola strain B14 resistant to rifampicin have also been isolated in a similar manner (Shoemaker et al., 1991). However, the rate of spontaneous mutagenesis is sufficiently low that without a strong selection strategy (e.g. antibiotic resistance) the isolation of mutants defective in certain enzymes or metabolic pathways is impractical. Although advances have been made in the development of gene transfer protocols for use with ruminal bacteria (for example, Shoemaker et al., 1991; Gardner et al., 1996), there are still limitations in terms of transfer frequency and available selective markers, to faciltiate the use of transposon mutagenesis or gene displacement strategies with these bacteria. For these reasons, we have successfully adapted the mutagenesis procedures described by Miller (1992) for use under anaerobic conditions and utilizing ethyl methane sulphonate (EMS) as a mutagen. Maximal mutation frequencies were achieved by exposure to EMS for periods of 30 min or longer, and were approximately 200-fold higher than the frequency of spontaneous mutation (Madeira et al., 1997). However, the EMS-induced frequency was still much lower than that typically observed with enteric bacteria, and we chose to also use an ampicillin enrichment procedure before screening for mutants (Wen, 1998). These mutagenesis procedures have not only allowed us to ask fundamental questions relating to bacterial physiology, but also to quantitatively evaluate the implications of changes/loss of enzyme activity on ruminal nitrogen metabolism. Our efforts to date have focused on the peptidase and glutamate dehydrogenase activities of P. bryantii. Wallace and co-workers have shown that the predominant forms of peptidase activity measurable in ruminal contents appear to be dipeptidyl aminopeptidases, most

The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism

105

active against neutral oligopeptides such as penta-alanine, as well as the synthetic substrate L-glycyl-L-argininyl-methylnapthylamide (Gly-Arg-MNA; Wallace and McKain, 1989; Wallace, 1994). Using pure cultures of ruminal bacteria, Wallace et al. (1990) and Wallace and McKain (1991) showed that P. ruminicola and B. fibrisolvens both exhibited a dipeptidyl aminopeptidase-like activity against penta-alanine, but P. ruminicola isolates were the only bacteria with measurable Gly-Arg-MNAse (hereafter referred to as PrtA) activity. McKain et al. (1992) extended these studies by isolating bacteria from ruminal contents and screening for PrtA activity and of the twelve positive isolates obtained, ten were consistent with the subgroupings of Prevotella spp. outlined by Avgustin et al. (1994). Wallace’s group has since attempted to use chromatographic procedures to enrich and purify the peptidases produced by Prevotella spp. Main peaks of activity following size exclusion chromatography of P. bryantii and Prevotella brevis extracts were estimated to contain proteins in excess of 100 kDa (Wallace et al., 1995); ion-exchange chromatography of Prevotella albensis cell extracts separated peptidase activity into four distinct peaks with different substrate specificities (Wallace et al., 1997). Madeira et al. (1997) found the inhibition profile of the PrtA was strikingly similar to that of gingipain, an extracellular, trypsin-like enzyme isolated from Porphyromonas (Bacteroides) gingivalis, that requires cysteine for activation and calcium for stabilization (Chen et al., 1992). Despite these efforts, none of these peptidases appear to have been purified to homogeneity and the relative contribution of individual enzymes to proteolysis and ammonium production by ruminal bacteria were not quantified. Within this context, Madeira et al. (1997) chose to generate mutants of P. bryantii strain B14 defective in PrtA activity for three primary reasons: (i) to assess the physiological role of this enzyme activity on growth of P. bryantii; (ii) to quantify the contribution from this enzyme to ruminal proteolysis, and ammonium production by the obligate aminoacid- and peptide-fermenting bacteria isolated by Russell and co-workers; and (iii) to facilitate comparative examination of wild type and mutant strains to identify polypeptide(s) responsible for this activity. Two independently derived mutants were isolated by Madeira et al. (1997) and in addition to the virtual elimination of measurable PrtA activity in both mutants, activity towards Arg-Arg-MNA was also lost. Interestingly, a second cysteine proteinase from P. gingivalis, termed argingipain, which presented a narrow specificity for synthetic substrates containing Arg in the P1 site and hydrophobic amino acids in the P2 and P3 sites, has also been described (Kadowaki et al., 1994) and subsequent studies showed gingipain and argingipain to be the same enzyme (Okamoto et al., 1995). Therefore, the P. bryantii mutant strains may lack Arg-ArgMNAse activity because the same enzyme is responsible for both activities, rather than the mutation(s) giving rise to polar or pleiotropic effects on the expression of multiple genes. That the mutation(s) in the mutant strains does not result in pleiotropic effects is further supported by the proteome profiles of the wild type and mutant strains, which are virtually indistinguishable from each other. The similarity of the results obtained with P. gingivalis, and P. bryantii, also suggests that this family of cysteine proteinases, although not identified in other eubacterial lines of descent, is widespread amongst bacteria belonging to the family Bacteroidacea. Although the physiological role of Gly-Arg-MNAse enzyme activity in P. bryantii is still not clear, an ecological role for this enzyme activity was clearly demonstrated by co-culture experiments with the obligate amino-acid-fermenting bacteria isolated by

106

M. Morrison

Russell’s group. Gelatin hydrolysate was chosen as the substrate for these experiments, to ensure that the results were directly comparable with earlier studies that established the prominent roles of C. aminophilum and P. anaerobius in controlling ruminal ammonia production (Chen and Russell, 1989). Both the rate and extent of ammonia production was decreased approximately 25% when the P. bryantii mutants replaced the wild type strain in co-cultures. Future strategies which specifically inhibit PrtA activity should therefore result in productive alterations in ruminal ammonia production. We also have isolated mutant strains of P. bryantii defective in their growth with peptides as sole N-source (Peng and Morrison, 1995) and Wallace’s group also have a set of mutant P. albensis strains which appear to possess reduced levels of all measurable peptidases. Further characterization of these mutants is currently underway and will probably provide further insights into the genetics and molecular biology controlling rate-limiting steps in proteolysis and peptide uptake in this numerically dominant group of ruminal bacteria. With such information, new productive strategies to control ruminal proteolysis should be forthcoming. We have also used these mutagenesis procedures to examine ammonia assimilation by P. bryantii. A number of Prevotella isolates have been found to lose cell viability following growth in a defined medium containing a relatively high concentration of glucose and a relatively low concentration of ammonia; this glucose toxicity was correlated with methylglyoxal production and its accumulation in the growth medium (Russell, 1992). More detailed studies with P. bryantii showed that, unlike enteric bacteria, methylglyoxal production occurred independently of phosphate limitation, perhaps instead related to aspects of carbon and/or nitrogen metabolism (Russell, 1993). Given that glutamate dehydrogenase (GDH) serves as a key enzyme linking carbon and nitrogen metabolism, we were interested in determining the role of this enzyme under different growth conditions. A mutant lacking GDH activity was isolated following EMS mutagenesis and ampicillin enrichment procedures (Z. Wen and M. Morrison, unpublished data) and nucleotide sequence analysis of the gdhA gene from the mutant strain confirmed a missense mutation had been created within one of the highly conserved motifs of the GdhA protein subunit. A clone of the mutant gdhA gene also did not complement the E. coli glutamate auxotroph used to successfully isolate clones of the wild type gdhA gene from P. bryantii. A number of growth experiments have been conducted with this mutant strain, and the only measurable phenotype associated with the GDH mutant strain was a slower growth rate when ammonium was used as the sole nitrogen source (108 min as compared with 75 min for the mutant and wild type strains, respectively). The mutant strain was also still capable of growth with both limiting (1 mM) and excess (10 mM) ammonium. The mutant strain was also incapable of growth in the presence of the glutamine synthetase inhibitor methiosulphoxamine, unless trypticase or a supplement of glutamine was provided. Although GDH activity appears relevant for optimal growth of P. bryantii, these results confirm the bacterium possesses more than one route of ammonium assimilation and glutamate biosynthesis, which also appears to be functionally independent of the prevailing ammonium concentration in the growth medium. In summation, mutagenesis strategies offer a valuable and productive means of examining bacterial physiology, and to evaluate the implications on ruminal function that may arise from the inhibition or elimination of specific enzyme activities. Although the strategy outlined here has been productive, other methods of mutational

The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism

107

analysis will need to be developed, primarily to facilitate gene identification. Although progess has been slow, advances in rumen bacterial genetics are being made (Flint 1997) including the successful use of gene displacement strategies (Brooker and McCarthy, 1997). Such strategies must be broadened in scope and application if further advances in our understanding, and opportunities for manipulating ruminal function, are to be achieved.

Differential display RT-PCR The completed microbial genome sequencing projects have revealed that as many as 30% of the presumptive coding sequences in prokaryote genomes have no current match in databases and, therefore, are of unidentified function (Hinton, 1997). Recent advances in molecular biology facilitate new opportunities to identify and ascribe function to these so-called ‘FUN’ genes. Both in vivo expression technology and DD RTPCR have permitted the identification of stress-, acid-, starvation- and colonization-inducible genes in pathogenic bacteria (Wong and McClelland, 1994; Valdivia and Falkow, 1997). While the former is applicable with bacteria for which relatively sophisticated genetic techniques are available, DD RT-PCR can be applied to microorganisms for which little or no techniques in bacterial genetics have been developed. However, there currently appears to have been limited application of this approach other than in the study of bacterial pathogenesis. We have begun to utilize DD RT-PCR as part of our studies of Ruminococcus albus, a Gram-positive anaerobe best recognized for its role in cellulose degradation in ruminant animals. Although our efforts to date are not directly related to nitrogen metabolism, our findings do reflect the potential power associated with this technique. The DD RT-PCR procedure used with R. albus RNA was essentially the same as described by Liang et al. (1993) except RT of the DNA-free total RNA was performed with random hexamers, and [␣-33P] dATP was used in PCR amplification of cDNA. Ten-mer oligonucleotides were used as primers for second DNA strand synthesis, and were chosen on the basis of limited sequence homology with the 16S rDNA gene of R. albus strain 8. To confirm that the RNA was not contaminated with DNA, PCR reactions were carried out with RT samples which contained no reverse transcriptase. Both the RT and PCR reactions were done in duplicate, and the cDNA products resolved on 6% (wt/vol) polyacrylamide gels under denaturing conditions. Putative differentially displayed bands were excised from polyacrylamide gels, and the DNA was reamplified using the same primer and PCR conditions. The DNA was ethanol precipitated and cloned into TA-cloning vectors and the nucleotide sequence of the cloned DNA was obtained using universal T7 and M13 reverse primers. The cloned DNA was then used as a template to produce [␣-32P] dCTP-labelled probes by random-primer labelling and northern blot analysis, to confirm differential expression of specific transcripts. Of the 18 putative dESTs identified in these initial experiments, six were confirmed to be differentially expressed. Alignments of these nucleotide sequences with other entries in the databases revealed significant homology with genes involved with protein export and solute uptake (Larson and Morrison, 2000). Our use of DD RT-PCR has contributed valuable information about the response of R. albus to biochemical and physical cues, and will facilitate our efforts to develop a

M. Morrison

108

comparative model with other cellulolytic and/or Gram-positive bacteria. We are currently adapting these procedures to examine differential gene expression by P. bryantii in response to nitrogen source and availability, and similar efforts are already underway in other labs with Ruminococcus flavefaciens. The challenge will be to ensure that other genetic technologies are also advanced, to ensure that these efforts extend beyond an exercise in gene identification and influence our understanding of rumen physiological ecology.

Nitrogen control in ruminal bacteria Because microbial protein is generally the predominant source of amino acids for ruminants, animal scientists have invested considerable effort in trying to determine the ‘optimal’ ruminal nitrogen requirements to maximize microbial protein synthesis. However, there are still large variations in bacterial growth rates and cell yields, even when animals are fed similar diets (Owens and Goetsch, 1986; Cotta and Russell, 1997). Since the biochemical examination of ammonium assimilation and nitrogen regulation by Smith and Hespell in the early 1980s, nitrogen control in ruminal microorganisms has received scant attention. Perhaps this is due to the widely held belief that the proteolytic enzymes of ruminal microorganisms are not tightly regulated, and ruminal ammonium concentrations are unlikely to be growth-limiting under most production scenarios (Wallace et al., 1998). However, as outlined above, Prevotella spp. support toxic schemes of carbon metabolism (methylglyoxal production) when an energy source is in great excess relative to ammonium (Russell 1992, 1993). Profound effects on bacterial yield are also seen if Prevotella spp. are cultivated first on peptides, then changed to a growth environment containing ammonium as sole nitrogen source. Russell (1983) found that a relatively high peptide concentration (15 g l⫺1) blocked ammonium assimilation in the bacterium now classified as P. bryantii. While this concentration of peptides might seem high, Russell (1983) calculated that less than 5% of the peptide nitrogen available was actually utilized. Therefore, the ‘effective concentration’ of peptide nitrogen that blocked ammonia assimilation may actually be much lower. Russell (1983) also found that P. bryantii demonstrated a prolonged lag in ammonia assimilation and cell growth, even though glucose fermentation was significant, upon transfer to a medium lacking peptides. These findings show some similarity to earlier observations made by Pittman and Bryant (1964) with P. ruminicola. It is also interesting to note that Cotta and Russell (1982) could not establish P. bryantii in continuous culture with a medium lacking peptides, presumably using an inoculum previously grown with peptides. It appears that peptide nitrogen not only blocks ammonium assimilation in Prevotella spp., but fermentation of energy proceeds uncoupled from protein synthesis (energy spilling), once peptides are depleted. Such a situation might arise in cattle fed low protein forages, where the rumen receives some protein nitrogen, but the microbiota is primarily reliant upon recycled, non-protein nitrogen sources. However, not all ruminal bacteria are capable of energy spilling, but are instead subject to rapid rates of death and lysis once ammonium is depleted. Fibrobacter succinogenes, a predominant cellulolytic bacterium, is one such example (Wells and Russell, 1995). The rate of autolysis appears to be modulated by the expression of substance(s) which inactivate autolysins, the expression of which appears to be influenced by nitrogen availability (Wells and Russell, 1996).

The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism

109

Therefore, while the conceptual focus of nitrogen-control in ruminal bacteria has often been in relation to shifts in assimilatory and biosynthetic pathways, fluctuations in nitrogen availability also result in periods of toxic schemes of carbon metabolism, energy spilling, and elevated rates of bacterial autolysis. With the exception of Russell and co-workers, there has been relatively little attention paid to how fluctuations in ruminal peptide and ammonia concentrations, as well as other physicochemical factors, might affect the balance between anabolic and catabolic pathways, and thereby influence microbial yield via cell lysis and/or death. For these reasons, the study of nitrogen control in ruminal bacteria needs to be given greater emphasis. Given their numerical predominance in the rumen, combined with their somewhat unusual characteristics in terms of nitrogen sources used for growth and response to nitrogen availability, we have decided to examine nitrogen control in Prevotella spp. Unlike most other prokaryotes, but similar to other members of the family Bacteroidaceae, NADPH-utilizing GDH activity is greatest in all Prevotella species tested following growth with low exogenous concentrations of ammonium, and the greatest relative change in GDH activity appears to be down-regulation in response to growth in the presence of peptides. Addition of trypticase to cultures of all three Prevotella spp. already growing on ammonium resulted in substantial reductions in NADPH-utilizing GDH activity, although the time required for the decrease in enzyme activity was different among strains. In strain B14, significant reductions were not observed until one doubling time after the addition of trypticase (72 min), consistent with regulation at the level of GDH synthesis. In contrast, NADPH-utilizing GDH specific activity decreased 60% and 90% in P. ruminicola strain 23 and P. brevis strain GA33, respectively, within 20 min of the addition of peptides, which is more consistent with the regulation of enzyme activity. Interestingly, the NADH-utilizing GDH specific activity in strains 23 and GA33 appeared to increase once peptides had been added to the growth medium, and the cells were allowed to grow for one doubling time. We subsequently confirmed by nondenaturing PAGE that P. brevis actually produces a second, NAD-dependent GDH in response to growth on peptides (Wen and Morrison, 1997). Although the underlying mechanisms of this nitrogen control remain to be determined in Prevotella spp., some clues have been provided from analogous studies conducted with Bacteroides thetaiotaomicron. Similar to P. brevis, this bacterium possesses two glutamate dehydrogenases, an NAD(P)H-utilizing enzyme (GdhA) as well as an NAD(H)-utilizing enzyme (GdhB) and their regulation in response to nitrogen source has been examined (Baggio and Morrison, 1996; W. Xing, L. Peng, L. Baggio, T. Chen and M. Morrison, unpublished data). In the wild type strain, GdhA activity was maximal when ammonium was provided as sole nitrogen source, and GdhB activity was only detectable when peptides were provided as the sole nitrogen source. To examine whether gdhA and/or gdhB gene expression was nitrogen regulated, a reporter gene shuttle vector was constructed, which contained a promoterless copy of a bifunctional xylosidase/arabinosidase originally cloned from Bacteroides ovatus (xyaA; W. Xing, L. Peng, L. Baggio, T. Chen and M. Morrison, unpublished data). In the wild type strain carrying a gdhA–xyaA fusion, xylosidase activity was maximal following growth in ammonium-limited medium, and decreased in response to the inclusion of peptides in the growth medium. In wild type strains carrying a gdhB–xyaA fusion, xylosidase activity was measurable only when peptides were available. Primer extension analysis and examination of the nucleotide sequence upstream of the ORF encoded by gdhA identified

M. Morrison

110

putative cis-acting sequences, and a motif similar to the consensus NtcA (nitrogen control protein A) binding motif, the global nitrogen regulatory system of cyanobacteria was identified. A role for this element in the regulation of gdhA expression in response to nitrogen source was confirmed by PCR-mediated deletion of the motif, and using the reporter gene vector described above. Despite the predicted changes in GdhA enzyme activity in response to nitrogen source, xylosidase activity remained low in those transconjugants containing the reporter gene fused with the mutagenized gdhA promoter. From these results, we conclude the NtcA-like motif is actually a cis-acting element, and that it facilitates a positive regulation of gdhA expression in response to ammonium. In cyanobacteria, the NtcA protein is required for the transcriptional activation of glnN, the structural gene encoding the glutamine synthetase (GSIII) enzyme subunit, which to date, has only been identified in the cyanobacteria, Bacteroides, Prevotella, Butyrivibrio and Ruminococcus . The NtcA protein has also recently been shown to repress gene expression in Anabaena PCC 7120, notably, the rbcLS operon, which encodes the large and small subunits of ribulose-1,5-bisphosphate carboxylase (Rubisco; Jiang et al., 1997). It appears that in cyanobacteria, the NtcA protein has a role to play in balancing rates of nitrogen and carbon assimilation. Although putative NtcA binding motifs can be identified upstream of various nitrogen assimilation genes from the ruminal bacteria mentioned above, further studies are required to confirm conservation of this regulatory protein, in addition to the glnN gene.

Concluding remarks and future perspectives The ruminal habitat has provided microbiologists with a rich bounty of information pertaining to microbial schemes of anaerobic metabolism, and the interactions among ruminal microorganisms and the host animal have been well described. In the past, much of the information has been derived by cultivation of a particular microorganism either axenically, or as part of a restrictive subset of the rumen microbial consortia. This research has led to a number of strategies which offer the potential to manipulate ruminal nitrogen metabolism, although most of these strategies involve either feed additives (e.g. monensin or other ionophores) or ‘protected’ supplements of amino acids, peptides or protein. Further options for the productive alteration of ruminal nitrogen metabolism are still desired, especially for grazing and forage-fed ruminants and/or intensive production scenarios where nitrogen loss into the environment is a societal concern. Molecular technologies now provide the potential to more completely examine a microbe’s response to its growth environment, through the rapidly emerging fields of molecular ecology, genomics and proteomics. In the future, how might these technologies assist with the overall goal of optimizing ruminant production? With reference to nitrogen metabolism, there appears to be some value associated with identifying precise way(s) to inhibit specific enzymes, such as the gingipain-like activity produced by Prevotella spp. Although current information is limited, these peptidases appear to be functionally and structurally distinctive, relative to mammalian and other microbial enzymes, which should facilitate inhibitor specificity. However, it is also possible that we may not be fully aware of the complete suite of ruminal proteolytic enzymes, or their origin. For this reason a proteomics-based approach also seems valuable: to iden-

The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism

111

tify predominant proteolytic activities in ruminal samples and by reverse genetics, ‘track’ down the origin of this activity. Second, can we selectively control populations of proteolytic microorganisms using bacteriophages or other antimicrobial compounds such as bacteriocins? Third, and in addition to taxonomic probes, can gene-specific probes be used to evaluate the physiological state of rumen microorganisms and the physicochemical conditions necessary to ensure optimal growth and yield? Fourth, in terms of nitrogen control of physiological processes, what are the ‘signals’ which ruminal microorganisms actually respond to: ammonium, peptides, cell density, or other factors? Genomics-based approaches such as DD RT-PCR, which aim to examine how a microbe responds to changes in its growth environment, offer the potential to identify novel physiological processes, and their impact on ruminal function. There will be no shortage of practical ideas for ruminal microbiologists to examine in the future and a plethora of useful approaches will be available. The greatest challenge facing ruminal microbiologists will probably be to identify and procure the financial resources necessary to execute their research. However, those research endeavours which offer the greatest potential to provide novel information concerning microbial ecology, genetics, and physiology also possess the greatest likelihood of being supported. Fortunately, acquisition of such fundamental information is also likely to maximize the probability of achieving productive alterations in nitrogen metabolism, and nitrogen retention in ruminant production systems.

References Attwood, G.T. and Klieve, A.V. (2000) Molecular techniques for monitoring bacteria and bacteriophage populations in the rumen. Proceedings of the 8th International Symposium on Microbial Ecology (in press). Attwood, G.T., Klieve, A.V., Ouwerkerk, D. and Patel, B.K. (1998) Ammonia-hyperproducing bacteria from New Zealand ruminants. Applied and Environmental Microbiology 64, 1796–1804. Avgustin, G., Wright, F. and Flint, H.J. (1994) Genetic diversity and phylogenetic relationships among strains of Prevotella (Bacteroides) ruminicola from the rumen. International Journal of Systematic Bacteriology 44, 246–255. Baggio, L. and Morrison, M. (1996) The NAD(P)H-utilizing glutamate dehydrogenase of Bacteroides thetaiotaomicron belongs to enzyme family I, and its activity is affected by transacting gene(s) positioned downstream of gdhA. Journal of Bacteriology 178, 7212–7220. Brooker, J.D. and McCarthy, D.M. (1997) Gene knockout of the intracellular amylase gene by homologous recombination in Streptococcus bovis. Current Microbiology 35, 133–138. Chen, G. and Russell, J.B. (1988) Fermentation of peptides and amino acids by a monensin-sensitive ruminal Peptostreptococcus. Applied and Environmental Microbiology 54, 2742–2749. Chen, G. and Russell, J.B. (1989) More monensin-sensitive, ammonia-producing bacteria from the rumen. Applied and Environmental Microbiology 55, 1052–1057. Chen, Z., Potempa, J., Polanowski, A., Wikstrom, M. and Travis, J. (1992) Purification and characterization of a 50-kDa cysteine proteinase (gingipain) from Porphyromonas gingivalis. Journal of Biological Chemistry 267, 18896–18901. Cotta, M.A. and Russell, J.B. (1982) Effect of peptides and amino acids on efficiency of rumen bacterial protein synthesis in continuous culture. Journal of Dairy Science 65, 226–234. Cotta, M.A. and Russell, J.B. (1997) Digestion of nitrogen in the rumen: a model for metabolism of nitrogen compounds in gastrointestinal environments. In: Mackie, R.I. and White, B.A. (eds) Gastrointestinal Ecosystems and Fermentations, vol. 1. Chapman & Hall, New York, pp. 380–423.

112

M. Morrison Egan, A.R. (1965) The fate and effects of duodenally infused casein and urea in sheep fed on a low-protein roughage. Australian Journal of Agricultural Research 16, 169–175. Eschenlauer, S.C., McEwan, N.R., Calza, R.E., Wallace, R.J., Onodera, R. and Newbold, C.J. (1998) Phylogenetic position and codon usage of two centrin genes from the rumen ciliate protozoan, Entodinium caudatum. FEMS Microbiology Letters 166, 147–154. Ffoulkes, D. and Leng, R.A. (1989) Dynamics of protozoa in the rumen of cattle. British Journal of Nutrition 59, 429–436. Flint, H.J. (1997) The rumen microbial ecosystem – some recent developments. Trends in Microbiology 5, 483–488. Gardner, R.G., Russell, J.B., Wilson, D.B., Wang, G.-R. and Shoemaker, N.B. (1996) Use of a modified Bacteroides–Prevotella shuttle vector to transfer a reconstructed b-1,4-D-endoglucanase gene into Bacteroides uniformis and Prevotella ruminicola B14. Applied and Environmental Microbiology 62, 196–202. Hinton, J.C.D. (1997) The Escherichia coli genome sequence: the end of an era or the start of the FUN? Molecular Microbiology 26, 417–422. Hungate, R.E. (1960) Symposium on selected topics in microbial ecology. 1. Microbial ecology of the rumen. Bacteriological Reviews 24, 353–364. Jiang, F., Mannervik, B. and Bergman, B. (1997) Evidence for redox regulation of the transcription factor NtcA, acting both as an activator and a repressor, in the cyanobacterium Anabaena PCC 7120. Biochemistry Journal 327, 513–517. Jouany, J.P., Demeyer, D.I. and Grain, J. (1988) Effect of defaunating the rumen. Animal Feed Science and Technology 21, 229–265. Kadowaki, T., Yoneda, M., Okamoto, K., Maeda, K. and Yamamoto, K. (1994) Purification and characterization of a novel arginine-specific cysteine proteinase (argingipain) involved in the pathogenesis of periodontal disease from the culture supernatant of Porphyromonas gingivalis. Journal of Biological Chemistry 269, 21371–21378. Kleive, A.V. and Swain, R.A. (1993) Estimation of ruminal bacteriophage numbers by pulsedfield electrophoresis and laser densitometry. Applied and Environmental Microbiology 59, 2299–2303. Klieve, A.V., Hudman, J.F. and Bauchop, T. (1989) Inducible bacteriophages from ruminal bacteria. Applied and Environmental Microbiology 55, 1630–1634. Klieve, A.V., Gregg, K. and Bauchop, T. (1991) Isolation and characterization of lytic bacteriophages from Bacteroides ruminicola subsp. brevis. Current Microbiology 23, 183–187. Klieve, A.V., Turner, A.F and Heck, G.L. (1998) Dietary influences on bacteriophage numbers in the rumen. Proceedings of the Australian Society of Animal Production 22, 341. Krause, D.O. and Russell, J.B. (1996) A rRNA approach for assessing the role of obligate amino acid fermenting bacteria in ruminal ammonia production. Applied and Environmental Microbiology 62, 815–821. Larson, M.A. and Morrison, M. (2000) Application of the differential display RT-PCR technique to examine conditional gene expression in Ruminococcus albus. Proceedings of the 8th International Symposium on Microbial Ecology (in press). Liang, P., Averboukh, L. and Pardee, A.B. (1993) Distribution and cloning of eukaryotic mRNAs by means of differential display: refinements and optimizations. Nucleic Acids Research 21, 3269–3275. Lockington, R.A., Attwood, G.T. and Brooker, J.D. (1988) Isolation and characterization of a temperate bacteriophage from the ruminal anaerobe Selenomonas ruminantium. Applied and Environmental Microbiology 54, 1575–1580. McKain, N., Wallace, R.J. and Watt, N.D. (1992) Selective isolation of bacteria with dipeptidyl aminopeptidase type I activity from the sheep rumen. FEMS Microbiology Letters 95, 169–174. Madeira, H.M.F., Peng, L. and Morrison, M. (1997) Biochemical and mutational analysis of a

The Microbial Ecology and Physiology of Ruminal Nitrogen Metabolism

113

gingipain-like peptidase activity from Prevotella ruminicola B14, and its role in ammonia production by ruminal bacteria. Applied and Environmental Microbiology 63, 670–675. Mendoza, G.D., Britton, R.A. and Stock, R.A. (1993) Influence of ruminal protozoa on site and extent of starch digestion and ruminal fermentation. Journal of Animal Science 71, 1572–1578. Miller, J.H. (1992) A Short Course in Bacterial Genetics. A Laboratory Manual and Handbook for Escherichia coli and Related Bacteria. Cold Spring Harbor Laboratory Press, New York, pp. 135–142. Morrison, M. and Mackie, R.I. (1996) Nitrogen metabolism by ruminal microorganisms: current understanding and future perspectives. Australian Journal of Agricultural Research 47, 227–246. Newbold, C.J., Wallace, R.J. and Watt, N.J. (1992) Properties of ionophore-resistant Bacteroides ruminicola enriched by cultivation in the presence of tetronasin. Journal of Applied Bacteriology 72, 65–70. Nolan, J.V. and Leng, R.A. (1972) Dynamic aspects of ammonia and urea metabolism in sheep. British Journal of Nutrition 27, 177–194. Okamoto, K., Misumi, Y., Kadowaki, T., Yoneda, M., Yamamoto, K. and Ikehara, Y. (1995) Structural characterization of argingipain, a novel arginine-specific cysteine proteinase as a major periodontal pathogeneic factor from Porphyromonas gingivalis. Archives of Biochemistry and Biophysics 316, 917–925. Owens, F.N. and Goetsch, A.L. (1986) Digesta passage and microbial protein synthesis. In: Milligan, L.P., Grovum, W.L. and Dobson, A. (eds) Control of Digestion and Metabolism in Ruminants. Prentice-Hall, Englewood Cliff, New Jersey, pp. 196–223. Paster, B.J., Russell, J.B., Yang, C.M.J., Chow, J.M., Woese, C.R. and Tanner, R. (1993) Phylogeny of ammonia-producing ruminal bacteria Peptostreptococcus anaerobius, Clostridium sticklandii, and Clostridium aminophilum sp. nov. International Journal of Systematic Bacteriology 43, 107–110. Peng, L. and Morrison, M. (1995) Mutagenesis of Prevotella ruminicola B14 and selection of mutants defective in peptide utilization. Abstracts of the 23rd Biennial Conference on Rumen Function. no. 26. Pittman, K.A. and Bryant, M.P. (1964) Peptides and other nitrogen sources for growth of Bacteroides ruminicola. Journal of Bacteriology 88, 401–410. Reilly, K. and Attwood, G.T. (1998) Detection of Clostridium proteoclasticum and closely related strains in the rumen by competitive PCR. Applied and Environmental Microbiology 64, 907–913. Russell, J.B. (1983) Fermentation of peptides by Bacteroides ruminicola B14. Applied and Environmental Microbiology 45, 1566–1574. Russell, J.B. (1992) Glucose toxicity and inability of Bacteroides ruminicola to regulate glucose transport and utilization. Applied and Environmental Microbiology 58, 2040–2045. Russell, J.B. (1993) Glucose toxicity in Prevotella ruminicola: methylglyoxal accumulation and its effects on membrane physiology. Applied and Environmental Microbiology 59, 2844–2850. Shoemaker, N.B., Anderson, K.L., Smithson, S.L., Wang, G.R. and Salyers, A.A. (1991) Conjugal transfer of a shuttle vector from the human colonic anaerobe Bacteroides uniformis to the ruminal anaerobe Prevotella (Bacteroides) ruminicola. Applied and Environmental Microbiology 57, 2114–2120. Swain, R.A., Nolan, J.V. and Klieve, A.V. (1996) Natural variability and diurnal fluctuations within the bacteriophage population of the rumen. Applied and Environmental Microbiology 62, 994–997. Valdivia, R.H. and Falkow, S. (1997) Fluorescence-based isolation of bacterial genes expressed within host cells. Science 277, 2007–2011. Virtanen, A.I. (1966) Milk production of cows on protein-free feed. Science 153, 1603–1614. Wallace, R.J. (1994) Amino acid and protein synthesis, turnover and breakdown by ruminal microorganisms. In: Asplund, J.M. (ed.) Principles of Protein Nutrition of Ruminants. CRC Press, Boca Raton, Florida, pp. 72–111.

114

M. Morrison Wallace, R.J. and McKain, N. (1989) Analysis of peptide metabolism by rumen microorganisms. Applied and Environmental Microbiology 55, 2372–2376. Wallace, R.J. and McKain, N. (1991) A survey of peptidase activity in rumen bacteria. Journal of General Microbiology 137, 2259–2264. Wallace, R.J., McKain, N. and Newbold, C.J. (1990) Metabolism of small peptides in rumen fluid. Accumulation of intermediates during hydrolysis of alanine oligomers, and comparison of peptidolytic activities of bacteria and protozoa. Journal of the Science of Food and Agriculture 50, 191–199. Wallace, R.J., Kopecny, J., Broderick, G.A., Walker, N.D., Sichao, L., Newbold, C.J. and McKain, N. (1995) Cleavage of di- and tri-peptides by Prevotella ruminicola. Anaerobe 1, 335–343. Wallace, R.J., McKain, N., Broderick, G.A., Rode, L.M., Walker, N.D., Newbold, C.J. and Kopecny, J. (1997) Peptidases of the rumen bacterium Prevotella ruminicola. Anaerobe 3, 35–42. Wallace, R.J., Onodera, R. and Cotta, M.A. (1998) Metabolism of nitrogen containing compounds. In: Hobson, P.N. (ed.) The Rumen Microbial Ecosystem, 2nd edn. Elsevier Applied Science, New York, pp. 283–328. Wallace, R.J., McKain, N., Miyagawa, E., Chaudary, L., McEwan, N.R. and Newbold, C.J. (1999) Eubacterium pyruvovorans, a new non-saccharolytic anaerobe from the rumen which ferments pyruvate and amino acids and carries out reverse ␤-oxidation of propionate and butyrate. Abstracts of the 99th Meeting of the American Society for Microbiology p. 381. Wells, J.E. and Russell, J.B. (1995) The effect of growth and starvation on the lysis of the ruminal cellulolytic bacterium Fibrobacter succinogenes. Applied and Environmental Microbiology 62, 1342–1346. Wells, J.E. and Russell, J.B. (1996) Why do ruminal bacteria die and lyse so quickly? Journal of Dairy Science 79, 1487–1495. Wen, Z. (1998) A molecular-based examination of the glutamate dehydrogenase and glutamine synthetase of ruminal Prevotella spp. PhD thesis, University of Nebraska, Nebraska, USA. Wen, Z. and Morrison, M. (1997) Glutamate dehydrogenase activity profiles for the type strains of ruminal Prevotella spp. Applied and Environmental Microbiology 63, 3314–3317. Wong, K.K. and McClelland, M. (1994) Stress-inducible gene of Salmonella typhimurium identified by arbitrarily primed PCR of RNA. Proceedings of the National Academy of Sciences USA 91, 639–643. Wright, A.D., Dehority, B.A. and Lynn, D.H. (1997) Phylogeny of the rumen ciliates Entodinium, Epidinium and Polyplastron (Litostomatea:Entodiniomorphida) inferred from small subunit ribosomal RNA sequences. Eukaryotic Microbiology 44, 61–67.

III

Nutrient Absorption and Splanchnic Metabolism

7

Tissue, Cellular and Molecular Aspects of Peptide Absorption and Utilization K.E. WEBB, JR Department of Animal and Poultry Sciences, Virginia Polytechnic Institute and State University, Blacksburg, Virginia, USA

Introduction As we enter the new millennium, we can reflect on recent efforts and know that an important body of literature, primarily from biomedical research, has accumulated over the last 10 years that is helping to define the physiological relevance of peptide absorption. No longer are we limited to just suggesting that peptide absorption occurs, now we are aware of the existence of special proteins that are responsible for transmembrane movement of peptides. The mRNA for some of these proteins has been cloned and the structural and functional characteristics of the encoded proteins are being determined. As the basis for understanding peptide absorption has grown, so also has the interest of animal scientists. Our laboratory has been conducting investigations in this area for over 20 years and now several others join us. In contrast with the biomedical community, our interest as animal scientists is to explain the appropriate dietary foundation of nutrient absorption that will result in the desired growth, development, and production in animals. Thus, we are interested in peptide absorption from the gastrointestinal tract, but our interests go beyond this. We are interested in knowing what, if any, role peptides may have as sources of amino acids for protein synthesis in different tissues and how they might be involved in controlling this process. These and other issues will be addressed in this paper. The primary focus will be on the ruminant, but some attention will be given to the human and laboratory species, simply because of the greater availability of information regarding these species.

Peptide transport: cellular and molecular aspects Functional characteristics of an ovine peptide transporter Absorption of Met-Gly and carnosine (␤-Ala-His) was observed to occur across ruminal and omasal epithelia (Matthews and Webb, 1995). At the substrate concentrations © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

117

118

K.E. Webb, Jr

tested, absorption appeared to be primarily non-mediated and the omasal epithelium appeared to have a greater capacity than ruminal epithelium to absorb and translocate these peptides. In a subsequent study, poly(A)+ RNA was extracted from omasal epithelium and injected into Xenopus laevis oocytes (Matthews et al., 1996b). Oocytes injected with sheep omasal epithelium poly(A)+ RNA absorbed more glycylsorcosine (Gly-Sar) than did oocytes injected with water. This expressed peptide transport was shown to be Na+ independent and pH dependent (e.g. maximal induced absorption of Gly-Sar occurred at pH 5.0–6.0). Gly-Sar uptake by oocytes injected with poly(A)+ RNA was shown to be saturable and have an affinity constant (Kt) of 0.40 ± 0.23 mM at pH 5.5. Uptake of 1 mM Gly-Sar by poly(A)+ RNA injected oocytes was inhibited by 5 mM ␤-Ala-His, Met-Gly, and Gly-Leu but not by 5 mM glycine, suggesting that these dipeptides were transported by the same expressed transport protein(s) but glycine was not. Oocytes injected with ovine poly(A)+ RNA and with a degenerate DNA oligomer that was complementary to the sequence flanking the initiation site of the cloned rabbit intestinal H+/peptide cotransport protein (Fei et al., 1994) completely inhibited Gly-Ser uptake. This indicated that there is some homology between the rabbit peptide transporter and the ovine peptide transporter expressed under these circumstances. In order to determine the structural features of peptides that might influence their affinity for the peptide transporter, X. laevis oocytes were injected with ovine poly(A)+ RNA and uptake at pH 5.5 of several peptides was examined (Pan et al., 1997). Oocytes injected with poly(A)+ RNA showed transport ability for the dipeptides GlySar, Gly-Leu, Gly-Pro, Phe-Leu and Leu-Leu. Oocytes showed transport ability for the tripeptides Leu-Ser-Phe, Leu-Gly-Phe, Lys-Tyr-Lys, Ala-Pro-Gly, Met-Leu-Phe and Leu-Leu-Tyr. The tetrapeptides, Met-Gly-Met-Met, Val-Gly-Asp-Glu, Ala-Gly-Ser-Glu and Val-Gly-Ser-Glu were transported but Pro-Phe-Gly-Lys and Val-Ala-Ala-Phe were not. No uptake occurred with the penta- (2), hexa- (1), septa- (1) and octapeptides (1) examined. These peptides constitute a variety of substrates varying in their molecular size, hydrophobicity and electrical charge under the experimental conditions used. For the di-, tri-, and tetrapeptides examined, the relationship between transport and molecular weight (MW) was r = 0.02 (P = 0.94), while the relationship between transport and hydrophobicity values was r = 0.34 (P = 0.21). When the relationship between transport and net charge was examined, values of r = 0.1 (P = 0.72) were obtained. Thus the transporter(s) seems to be capable of translocating substrates regardless of their MW, hydrophobicity or charge. Recent studies from other laboratories on the mammalian proton-coupled peptide transporter, PepT1, also demonstrate that peptide substrates are transported by electrogenic, H+-coupled cotransport, which is independent of their physicochemical characteristics (i.e. size, charge; Daniel et al., 1996; Wenzel et al., 1996). When X. laevis oocytes were injected with cRNA for ovine PepT1 (Pan, 1999), peptide uptake by the oocytes was observed to be pH and concentration dependent. External pH was varied by bathing the oocytes in buffers ranging from pH 5.0 to 7.0. Uptake of Glu-Glu and Met-Met was greatest at pH 5.0, uptake of Gly-Sar was greatest at pH 5.0–6.0, and uptake of Lys-Lys was greatest at pH 7.0. These results agree with previously published reports indicating that an inwardly directed H+ gradient as well as the charge on the peptide both influence transport (Fei et al., 1994). The Kt for Gly-Sar uptake at pH 5.5 was observed to be 0.61 mM which is similar to the Kt of

Tissue, Cellular and Molecular Aspects of Peptide Absorption and Utilization

119

0.40 mM previously observed in our laboratory (Matthews et al., 1996b). The Kts for Met-Met, Lys-Lys and Glu-Glu uptake at pH 5.5 were 27.5, 20.4 and 28.7 ␮M, respectively. Affinity constants ranging from 81 ␮M for Gly-Leu to 2.5 mM for GlyGly were reported by Fei et al. (1994). Peptide absorption across ruminal and omasal epithelia has been demonstrated in vitro as has mRNA in sheep omasal epithelial cells that encodes for a peptide transporter(s). Substrate specificity of the transporter(s) indicates that many but not all di-, tri- and tetrapeptides can be transported. Initial kinetic evaluation of ovine PepT1 indicates a range in affinity constants for substrates. While results from our initial research indicated that absorption may be largely by non-mediated processes, it is obvious that mechanisms for mediated transport of peptides in the ruminant stomach are present. Omasal epithelium was collected from sheep and mounted in parabiotic chambers to measure the uptake of Gly-Sar in the presence of other peptides (McCollum and Webb, 1998). If multiple substrates are transported on a given transporter, as peptides are believed to be, then they compete with one another if more than one is present. To our surprise, co-incubating high levels of potentially competing peptide substrates with Gly-Sar resulted in a stimulation and not a depression in Gly-Sar uptake. Even though our previous results clearly indicate that mRNA for a peptide transporter is present in omasal epithelium, these results are certainly not consistent with mediated transport being the only mechanism involved in peptide transport across omasal epithelium. Madara and Pappenheimer (1987) discussed paracellular transport through intestinal epithelia and showed that a prerequisite for this process is the mediated uptake of the substrate. It may be that paracellular transport is the mechanism that was responsible for what appeared to be a nonmediated uptake of peptides by epithelial tissues mounted in parabiotic chambers (Matthews and Webb, 1995). From the evidence that is accumulating, it seems reasonable to suggest that multiple mechanisms may be involved in the transport of peptides across gastrointestinal epithelia. Validation of the presence of and the clarification of the relative importance of each of these mechanisms will increase our understanding of peptide absorption and the contribution this may make to the overall nutritional status of the animal.

Tissue distribution of mRNA for an ovine and bovine peptide transporter Sheep omasal epithelium poly(A)+ RNA was used as the template for reverse transcriptase PCR (RT-PCR) to develop a probe for Northern blot analysis (Chen et al., 1999). The published sequences of rabbit, human and rat PepT1 (Fei et al., 1994; Liang et al., 1995; Miyamoto et al., 1996) were used to design oligonucleotides to a highly homologous region (from transmembrane domain III to VI). After RT-PCR amplification, a 446-bp fragment from sheep omasal epithelium was cloned and sequenced. The predicted amino acid sequence of this fragment was 85.8%, 90.5% and 90.5% identical to rabbit, human and rat PepT1, respectively. The fragment has low similarity with any known PepT2 sequences. Poly(A)+ RNA isolated from the rumen, omasum and small intestine of sheep and lactating dairy cows showed positive hybridization to the probe in Northern blot analysis. The size of the mRNA that hybridized to the probe was 2.8 kb for both sheep and dairy cows. From the summary of the Northern blot analysis for

K.E. Webb, Jr

120

the five sheep and three cows tested, the pattern of distribution appeared to be consistent within the same species, although the relative abundance varied between animals. Among these tissues, the abundance of PepT1 was higher in the jejunum and ileum in comparison with omasum and duodenum, while abundance was lowest in the rumen. PepT1 was not detectable in the abomasum, liver, kidney, caecum, colon, longissimus and semitendinosus muscles in either sheep or cows, or in the mammary gland in cows.

Structure of an ovine peptide transporter Recently, we have cloned a full-length cDNA for sheep PepT1 (Pan, 1999). A partial sequence of this clone indicates that it has about 85% amino acid identity with human PepT1 in the non-loop regions. Amino acid identity within the large extracellular loop appears to be much lower. The existence of poly(A)+ RNA transcript(s) is strong evidence for the presence of a peptide transport protein(s). Thus, peptide absorption from the stomach and small intestine of ruminants appears to be a physiologically relevant process.

Overview of human, rabbit and rat peptide transporters The importance of peptide transport has been documented in recent years. By taking advantage of molecular technology, a few peptide transporters have been identified. By cloning and in vitro expression, the molecular structures of these peptide transporters and the mechanisms by which the systems work have begun to be revealed. It was not until the mid 1990s that the structures of the peptide transporters became clear. In 1994, two groups simultaneously reported the identification of a rabbit oligopeptide transporter designated PepT1 (Boll et al., 1994; Fei et al., 1994). This was the beginning of a new era of peptide transport studies. The reported size of rabbit PepT1 cDNA was 2.7 kb (Fei et al., 1994). It codes for a protein consisting of 707 amino acid residues. The amino acid sequence indicates that PepT1 is a membrane protein with 12 membrane-spanning domains. The protein has a large hydrophilic loop between domains 9 and 10 and this makes the protein different from transporters reported previously. Based on the information from in vitro translation, this loop is the target for N-linked glycosylation and the authors believe it to be an extracellular portion of the protein. Other features, including a protein kinase C site and a cAMPdependent phosphorylation site, appear also to be present on the protein. Others have cloned human (Liang et al., 1995) and rat (Saito et al., 1995) PepT1. Human PepT1 cDNA is 2.2 kb long with an open reading frame encoding for a protein composed of 708 amino acids (Liang et al., 1995). Rat PepT1 cDNA is 2.9–3.0 kb long with an open reading frame encoding a 710 amino acid protein (Saito et al., 1995; Miyamoto et al., 1996). Despite the different sizes of the PepT1 cDNA of these species obtained by the different groups, they all encode proteins that are very similar in structure. The 12 putative transmembrane domains and the long hydrophilic segment between domains 9 and 10 are the most significant features for this group of transport proteins. The pre-

Tissue, Cellular and Molecular Aspects of Peptide Absorption and Utilization

121

dicted structural model shows a long hydrophilic loop on the extracellular side containing several N-linked glycosylation sites. With all the 12 ␣-helices accommodated within the membrane, both the amino and carboxyl termini are located on the cytoplasmic side. These PepT1 proteins are all highly homologous. The amino acid sequence of the rat PepT1 is 77% and 83% identical with the rabbit and human PepT1, respectively (Saito et al., 1995). The rabbit and human PepT1 are 81% identical (Liang et al., 1995). The amino acid sequences of these transporters do not show strong homology with other known classes of transport proteins. It is worth mentioning that, while all of the transmembrane domains are highly conserved, the extracellular loops are much less so. Structural differences among these transport proteins indicate that there are different numbers and locations of sites for protein kinase A and C phosphorylation. The extent to which these differences may affect the function of these transporters is still not clear. Another peptide transporter (PepT2) that is structurally and functionally different from PepT1 has been identified (Liu et al., 1995; Boll et al., 1996; Saito et al., 1996). The human full-length PepT2 cDNA is 2.7 kb with an open reading frame of 2.2 kb (Liu et al., 1995). The rat PepT2 cDNA is 3.9 kb (Saito et al., 1996). Although the sizes of these PepT2 cDNAs are different, they both encode a 729 amino acid protein (Boll et al., 1996). The shared features include the putative 12 transmembrane domains, the large extracellular loop, and several sites for protein kinase dependent phosphorylation. These structural features are observed in both the PepT1 and PepT2 transporter groups. Even with these similarities, these appear to belong to two distinct transporter groups. When PepT1 and PepT2 are compared from the same species, they display lower identity than do the members in the same group. For example, human PepT2 has 83% amino acid identity with rat PepT2 whereas only 50% with human PepT1. PepT2 proteins have a greater molecular mass than PepT1. Transport activities of these cloned peptide transporters have been characterized mainly by in vitro expression of the cDNA in X. laevis oocytes or other cell lines. Transport studies have shown that the cloned peptide transporters are capable of taking up a broad range of di- and tripeptides, regardless of whether they contain acidic, basic or hydrophobic amino acids. The peptides transported may be in either the charged or neutral form under different conditions. Results from studies designed to examine the effects of the net charge of a substrate on peptide transport suggest that neutral substrates are preferred by the peptide transporter when compared with charged peptides under physiological pH conditions (Lister et al., 1997). Amasheh et al. (1997) suggest that, under physiological conditions, the affinity of PepT1 for zwitterionic or anionic substrates is greater than for cationic substrates. For transport of a cationic substrate, the pH must be higher. One of the features of peptide transporters is the necessity for the proteins to cotransport a proton along with the peptide substrate. Brandsch et al. (1997) studied the effect of protons on the affinity and Vmax of Gly-Sar uptake by Caco-2 (PepT1) and SKPT cells (PepT2). Uptake of Gly-Sar in both cells was measured over a concentration range at an extracellular pH of either 6.0 or 7.0. In Caco-2 cells, the Kt was 1.0 mM at pH 6.0 and 1.2 mM at pH 7.0. However, the Vmax was 13.7 ± 0.3 nmol 10 min⫺1 mg⫺1 of protein at pH 6.0 and 5.8 ± 0.3 nmol 10min⫺1 mg⫺1 of protein at pH 7.0. Similar results were obtained with SKPT cells. In both cell types, protons affected only Vmax but did not affect the affinity of the transporters for the substrates. Steel et al.

122

K.E. Webb, Jr

(1997) suggested that there is a proton–peptide coupling ratio of 1:1, 2:1 and 1:1 for neutral, acidic and basic dipeptides, respectively. They also showed that, at a pH of 5.5–6.0, PepT1 favoured substrates in neutral and acidic forms. Given the role of these proteins as peptide transporters, factors that influence the protein structure, its localization on the membrane, and the specific position for substrate binding may all have an effect on the transport activity. It was reported that, among histidyl residues present, His-57 in PepT1 and His-87 in PepT2 are the most critical histidyl residues necessary for the uptake function and probably represent critical binding sites (Fei et al., 1997). It may be that protein kinases C and A are involved in the regulation of peptide transport. Muller et al. (1996) showed that increased intracellular levels of cAMP in Caco-2 cells that expressed PepT1 resulted in a 50% reduction of Gly-Sar uptake. Zinc is an essential trace element that plays a fundamental role in the structure and function of many proteins, e.g. stabilizing the structure of an enzyme, being an essential component of the active site of an enzyme, and as a regulatory factor. Daniel and Adibi (1995) concluded that zinc had a selective effect on peptide transport. They incubated brushborder membrane vesicles (BBMV) with zinc sulphate and observed an increased uptake of Gly-Gln and Leu-Tyr, without changing the diffusion rate of the substrates. Zinc had no effect on the uptake of either Gln or glucose by BBMV. Information on the distribution of the peptide transporters among various tissues comes largely from the search for the messenger RNA that encodes for the protein. Along with the cloning of PepT1 and PepT2 from various species, distribution of their mRNA has been studied. The basic technique is Northern blot analysis using a specific radiolabelled DNA probe, where for most studies reported, full-length cDNA was used. A 2.9 kb mRNA was found in the small intestine of the rabbit, while much lower levels were observed in the liver and kidney and only trace amounts were found in the brain (Fei et al., 1994). No mRNA was detected in the colon, skeletal muscle, heart, spleen or lung. The mRNA for PepT1 was 2.9–3.0 kb in the rat, the major location of which was in the small intestine (Saito et al., 1995; Miyamoto et al., 1996). Trace amounts of mRNA were found in kidney cortex, but none was observed in liver. Liang et al. (1995) reported the size of human PepT1 mRNA to be 3.3 kb, based on Northern blot analysis. They observed a major presence of this PepT1 mRNA in the small intestine as well as in an intestinal epithelium derived cell line (Caco-2). They also observed PepT1 mRNA in kidney, placenta, liver and pancreas. The mRNA transcript was absent in muscle, brain and heart. A ~4 kb mRNA for PepT2 was found mainly in the kidney medulla and at lower abundance in the kidney cortex (Saito et al., 1996). The mRNA was also detected in brain, lung and spleen, but was undetectable in the heart, liver and small intestine. In the rabbit, a 4.8 kb mRNA was found in the kidney cortex as well as in brain, lung, liver and heart (Boll et al., 1996). PepT2 mRNA was detected in human kidney and small intestine (Liu et al., 1995). PepT1 mRNA was expressed all along the small intestine and to a much lesser extent in the colon (Freeman et al., 1995). PepT1 mRNA was not detected in the stomach, sacculus rotundus or caecum. Expression was restricted to the epithelial surface of the small intestine and there was no detectable expression in deeper tissues such as the lamina propria, muscularis mucosa or serosa (Fei et al., 1994). Along the crypt–villus axis, the mRNA was detected at or above the crypt–villus junction with the maximal

Tissue, Cellular and Molecular Aspects of Peptide Absorption and Utilization

123

expression occurring at about 100–200 ␮m above the junction. The mRNA was absent in the lower-to-mid crypt throughout the entire small intestine. Ogihara et al. (1996) were the first to investigate the localization of the transporter protein itself instead of its mRNA. They developed an anti-PepT1 antibody and used it (by means of immunoblotting) to look for PepT1 protein in the rat. Their results confirmed the exclusive expression of PepT1 throughout the length of the small intestine and the absence of PepT1 in the crypt. They also found that PepT1 was specific to the differentiated absorptive epithelial cells and was located mainly on the brush-border membrane of the cell. From these results, we can conclude that PepT1 and PepT2 are distributed differently between the tissues. PepT1 is mainly an intestinal peptide transporter whereas PepT2 is mainly a renal peptide transporter. Interestingly, PepT1 is also detectable in the kidney. The existence of PepT1 in the small intestine of animals suggests its nutritional importance. The PepT2 in the kidney may play a significant role in conserving peptide forms of amino nitrogen by means of reabsorption (Daniel and Herget, 1997).

Peptide absorption In vitro characterization of gastrointestinal absorption of peptides Casein, soybean meal and distillers’ dried grains were incubated in a buffered ruminal fluid inoculum for 8 h (Jayawardena, 2000). Following incubation, cell-free supernatants were obtained by centrifugation and these were used as the mucosal fluids in parabiotic chambers containing either ruminal or omasal epithelium. Initially, free amino acid concentrations ranged from 7.2 to 60 mg l⫺1 and peptide amino acid concentrations ranged from 100 to 270 mg l⫺1 in these mucosal buffers. Serosal appearance of free and peptide amino acids was measured after 240 min. For ruminal epithelium, serosal appearance of free amino acids from casein, soybean meal, and distillers’ dried grains was 242, 220 and 234 mg l⫺1 mg⫺1 dry tissue, respectively. Corresponding figures for peptide amino acids were 493, 329 and 453 mg l⫺1 mg⫺1 dry tissue. For omasal epithelium, serosal appearance of free amino acids from casein, soybean meal and distillers’ dried grains was 438, 323 and 340 mg l⫺1 mg⫺1 dry tissue. Corresponding figures for peptide amino acids were 2249, 807 and 1191 mg l⫺1 mg⫺1 dry tissue. Serosal appearance of peptide amino acids was greater than serosal appearance of free amino acids in both tissues, probably reflecting the concentration effect of the substrates in the mucosal buffer. Movement through omasal epithelium was much greater than through ruminal epithelium, especially for peptides. Uptake of Gly-Sar by sheep jejunal and ileal BBMV in a study we conducted showed that these membranes have the capability of translocating this dipeptide (Bowers, 1997). Uptake was greater in BBMV from jejunal tissue than from ileal tissue. Uptake of 0.3 mM Gly-Sar was not stimulated by an inwardly directed H+ gradient (pH 6.4 outside, pH 7.5 inside) in either jejunal or ileal BBMV. Uptake of 10 ␮M Gly-Pro by BBMV prepared from sheep duodenal epithelium was reported by Backwell et al. (1995), who found that uptake was dependent on a H+ gradient (based on results obtained from an intravesicular pH of 8.4 and extravesicular pHs of 6.0 and 8.4). At first glance, the results of these studies appear to be in conflict.

K.E. Webb, Jr

124

A careful examination of the protocol reported by Backwell et al. (1995) reveals that the extravesicular pH in their study was probably closer to 7.0 than to the pH 6.0 reported. Furthermore, an extravesicular pH of 8.4 would cause peptide substrates to be more negatively charged than would occur physiologically. Because peptide molecules need to be in zwitterionic form in order to be transported (Ganapathy and Leibach, 1985), the results reported by Backwell et al. (1995) may be due to a reduction in uptake because of the charged nature of the peptide substrate at pH 8.4 rather than an enhancement of uptake in the presence of a H+ gradient. BBMV prepared from the proximal jejunum of dairy cows was used to examine Gly-Sar uptake (Wolffram et al., 1998). Results from this study clearly indicate that an inwardly directed H+ gradient stimulated a greater rate of uptake of Gly-Sar. In contrast to the two previous studies (Backwell et al., 1995; Bowers, 1997), this transport was uphill. There is no clear explanation for this concentrative uptake of a peptide. The authors attributed their ability to measure concentrative uptake to the fact that they used a low substrate concentration (25 ␮M) while others used much higher concentrations (0.1–0.8 mM). Even though Backwell et al. (1995) used a concentration of 10 ␮M and did not observe concentrative uptake, there may be other problems with the design of this study, as previously discussed, that preclude observing a concentrative uptake. Wolffram et al. (1998) argue that if a transport mechanism with a low substrate affinity or transport capacity or both is responsible for peptide transport, then, under conditions of a high substrate concentration, the driving forces for uphill transport might be exhausted well before significant amounts of the substrate accumulates inside the vesicle.

Portal flux of peptides One of the more controversial issues regarding peptide transport revolves around the issue of the quantitative aspects of portal flux of peptides. Our suggestion many years ago that there may be a sizeable flux of peptides across portal-drained viscera (PDV) is the origin of this controversy. The fact that we suggested that the ruminant stomach may be involved in this process served only to heighten the controversy. This single issue, more than any other, has probably been responsible for stimulating laboratories other than ours to initiate investigations into peptide transport in ruminants. These aspects have been extensively reviewed previously by us (Matthews et al., 1996a; Webb, 1986, 1990; Webb and Bergman, 1991; Webb et al., 1992, 1993; Webb and Matthews, 1994, 1998). In the first study we conducted, we attempted to quantify the flux of free and peptide amino acids across the PDV of calves (Koeln et al., 1993). Quantifying peptides as the difference between total amino acids in the protein-free filtrate (following acid hydrolysis) and free amino acids (following sulphosalicylic acid precipitation), we observed that there was a much greater flux of peptide amino acids than free amino acids across the PDV. Subsequently, we reported that not only did peptides constitute a major fraction of PDV flux of amino acids but that non-mesenteric tissues contributed substantially to this (Webb et al., 1992, 1993). Seal and Parker (1996) reported that there was a considerable net appearance of peptide-N in both mesenteric-drained viscera (MDV) and PDV of steers. Net appear-

Tissue, Cellular and Molecular Aspects of Peptide Absorption and Utilization

125

ance of free amino acids in blood was greater than peptide amino acids across MDV than across PDV. Conversely, net appearance of peptide amino acids in blood was greater than free amino acids across PDV than across MDV. This is consistent with the ruminant stomach contributing significantly to the appearance of peptides in blood. Gastrointestinal absorption of peptides was examined in sheep fed a forage-based diet (Backwell et al., 1997). They failed to observe any net flux of peptide amino acids across either MDV or PDV. They also observed lower peptide concentrations in blood plasma than reported previously by Koeln et al. (1993) and Seal and Parker (1996). They attributed this absence of a net flux of peptide amino acids across MDV and PDV as being due to a matter of technique, more specifically a more efficient deproteinization of blood plasma, thus not creating artificially high ‘peptide’ concentrations. Lactating dairy cows were fed diets based on either dry-rolled or steam-flaked sorghum and portal fluxes of free and peptide amino acids were quantified (H. Tagari et al., personal communication). As would be expected, there was a positive portal flux for all amino acids. The quantity of peptide amino acids appearing in the portal vein was lower than free amino acids, but appearance was positive for 70% of the amino acids. This included fluxes differing (P < 0.059) from zero for Ala, Arg, Asp, Glu, Gly, His, Leu, Lys, Met, Ser, Thr and Val. The deproteinization procedure used in this study involved protein precipitation with methanol followed by ultrafiltration through 3000 mol. wt filters. Residual peptides would, therefore, be small. Evidence suggests a net appearance of peptide amino acids in portal and possibly mesenteric blood plasma. Methodological differences are likely contributors to variations observed in the magnitude of portal appearance of peptide amino acids. The procedure employed by H. Tagari et al. (personal communication) appears to be the soundest yet employed. Continued efforts in this area will provide further clarification regarding the absolute magnitude of the contribution of peptide amino acids to amino acid flux across MDV and PDV.

Utilization of circulating peptides Peptide utilization by cultured myogenic cells L-Methionine-containing

peptides were evaluated as sources of methionine to support protein accretion in C2C12 myogenic cells from mouse muscle (Pan et al., 1996). Expressed as a percentage of the response to free methionine, growth of C2C12 cells differed due to the type of dipeptide (11–108%). Met-Met, Met-Val, and Leu-Met were utilized as efficiently as free methionine. Pro-Met and Gly-Met were poorly utilized by C2C12 cells. Met-Pro, Phe-Met, Met-Phe, Met-Leu, Met-Gly, Ala-Met and Met-Ala were utilized at a rate of about 62–86% of the rate of free methionine and Met-Ser, Ser-Met and Val-Met were utilized at 26–43% of the rate of free methionine. Primary cultures of ovine myogenic satellite cells were evaluated for their ability to use peptide-bound methionine as a source of methionine for protein accretion and cell proliferation after isolation from skeletal muscle (Pan and Webb, 1998). The cultured myogenic cells were able to utilize all the methionine-containing dipeptides tested for protein accretion with responses ranging from about 49 to 95% of the response for free methionine. This is consistent with the concept that peptide-bound amino acids can

K.E. Webb, Jr

126

serve as amino acid sources for protein accretion in sheep skeletal muscle. In some cases, the molecular arrangement of the dipeptides with the same amino acid composition influenced the relative ability of the dipeptides to serve as methionine sources. For all peptides studied, however, only Ala-Met was utilized to support protein accretion as well as free methionine.

Peptide utilization by cultured mammary cells L-Methionine-containing

peptides were evaluated for their ability to be a source of methionine to support protein accretion in MAC-T, bovine mammary epithelial, cells (Pan et al., 1996). All of the methionine-containing dipeptides examined were able to support protein accretion in cultured MAC-T cells with the response ranging from 35% to 122% of the free methionine growth response. Met-Val, Met-Leu, Met-Met and Leu-Met supported greater protein accretion than did free methionine. Phe-Met, Met-Phe, Ala-Met, Met-Ala, Met-Ser and Met-Gly were utilized as effectively as free methionine. Gly-Met, Pro-Met and Ser-Met were the least utilized peptides in MAC-T cells. Dipeptides with methionine at the N-terminus were preferred substrates to those with methionine at the C-terminus. Regulation of the use of peptides as amino acid substrates for protein accretion and cell proliferation in MAC-T cells by serum factors was examined (Pan et al., 1998). Results indicated that adult animal sera from humans, horses, chickens, pigs and rabbits promote the utilization of most methionine-containing peptides. By themselves, neither insulin nor serum lipids were able to facilitate peptide utilization. Cultured MAC-T bovine mammary epithelial cells were used to study the ability of methionine-containing peptides to substitute for free methionine in the synthesis of secreted proteins (Wang, 1994). All of the methionyl peptides examined were utilized by the MAC-T cells as sources of methionine for the synthesis of both secreted and cellular proteins. Most of the methionine-containing peptides were as efficient as free methionine in promoting protein synthesis. These results indicate that MAC-T mammary epithelial cells are able to utilize small methionine-containing peptides as sources of methionine to support cellular protein accretion and the synthesis of secreted proteins.

Peptide utilization by mammary tissue explants Mammary tissue explants from lactating (10–11 days) CD-1 mice were used to study the ability of methionine-containing peptides on the synthesis of secreted proteins (Wang et al., 1996). Mammary tissue explants were able to utilize methionine from all peptides studied. Eleven of the peptides promoted 15–76% greater synthesis of secreted proteins than did free methionine. Dipeptides containing either valine or serine promoted the greatest synthesis. The remaining six peptides were not different from free methionine in promoting synthesis of secreted proteins. Mammary explants were able to utilize lysine from all lysyl peptides examined for the synthesis of secreted proteins (Wang, 1994). These lysyl peptides generally were similar to free lysine in promoting synthesis of secreted proteins. The synthesis of pro-

Tissue, Cellular and Molecular Aspects of Peptide Absorption and Utilization

127

teins promoted by Gly-His-Lys was about 17% greater than that promoted by lysine. The other peptides were not different from lysine in promoting protein synthesis which ranged from 91 to 108% of the synthesis promoted by lysine. Within each of the three peptide pairs, Asp-Lys and Lys-Asp, Gly-Lys and Lys-Gly, and Val-Lys and Lys-Val, location of the lysyl residue at either the N- or C-terminal position did not affect protein synthesis. These results are consistent with those previously discussed for cultured MAC-T mammary epithelial cells. Together, these studies indicate that a wide range of peptidebound methionine and lysine substrates can support the synthesis of milk proteins by mammalian epithelial cells, at least as well as free methionine and lysine.

Peptide flux across the mammary gland in vivo Use of free and peptide amino acids by the mammary gland of lactating goats was investigated by Backwell et al. (1994). They observed that Phe from Gly-Phe and Leu from Gly-Leu were incorporated into casein in the mammary gland. It is obvious that the incorporation of Phe and Leu from these peptides into casein was an intracellular process. The authors were unable to determine whether there was transport of intact peptides into the cell or whether extracellular hydrolysis preceded transport of the amino acids. Shennan et al. (1998) concur that amino acids of peptide origin are utilized by the mammary gland (rat), but their data indicate that peptides are probably not transported but are hydrolysed prior to the constituent amino acids entering the cell. We cannot rule out the possibility that a peptide transporter other than PepT1 may be involved, but data from our laboratory supports their claims because we were unable to detect mRNA for PepT1 in the mammary gland of lactating cows (Chen et al., 1999). Estimates of the extent of incorporation of Phe, Met, Lys and Tyr arriving at the mammary gland of lactating goats in peptide form have been recently reported (Bequette et al., 1999). They report that 5–11% of Phe, 8–18% of Met, 4–13% of Lys, and 13–25% of Tyr incorporated into casein is of peptide origin. This confirms earlier work from the same laboratory indicating that peptides can contribute amino acids for milk protein synthesis (Backwell et al., 1996). When fluxes of free amino acids across the mammary gland of lactating cows fed either steam-flaked or dry-rolled sorghum were examined, it was observed that, except for Gly, there was a net uptake of free amino acids (H. Tagari et al., personal communication). Uptake of amino acids of peptide origin by the mammary gland was much smaller but was positive for about 65% of the amino acids including positive fluxes that differed (P < 0.05) from zero for Glu, Leu, Lys and Ser. Collectively, these studies indicate that peptides contribute their constituent amino acids for the synthesis of milk proteins. In magnitude, these amino acids account for only a small portion of the total amino acids incorporated into the proteins. This source, however, may be critical and may play a role in controlling milk protein synthesis.

K.E. Webb, Jr

128

References Amasheh, S., Wenzel, U., Boll, M., Dorn, D., Weber, W.-M., Clauss, W. and Daniel, H. (1997) Transport of charged dipeptides by the intestinal H+/peptide symporter PepT1 expressed in Xenopus laevis oocytes. Journal of Membrane Biology 155, 247–256. Backwell, F.R.C., Bequette, B.J., Wilson, D., Calder, A.G., Metcalf, J.A., Wray-Cahen, D., MacRae, J.C., Beever, D.E. and Lobley, G.E. (1994) Utilization of dipeptides by the caprine mammary gland for milk protein synthesis. American Journal of Physiology 267, R1–R6. Backwell, F.R.C., Wilson, D. and Schweizer, A. (1995) Evidence for a glycyl-proline transport system in ovine enterocyte brush-border membrane vesicles. Biochemical and Biophysical Research Communications 215, 561–565. Backwell, F.R.C., Bequette, B.J., Wilson, D., Metcalf, J.A., Franklin, M.F., Beever, D.E., Lobley, G.E. and MacRae, J.C. (1996) Evidence for the utilization of peptides for milk protein synthesis in the lactating dairy goat in vivo. American Journal of Physiology 271, R955–R960. Backwell, F.R.C., Hipolito-Reis, M., Wilson, D., Bruce, L.A., Buchan, V. and MacRae, J.C. (1997) Quantification of circulating peptides and assessment of peptide uptake across the gastrointestinal tract of sheep. Journal of Animal Science 75, 3315–3322. Bequette, B.J., Backwell, F.R.C., Kyle, C.E., Calder, A.G., Buchan, V., Crompton, L.A., France, J. and MacRae, J.C. (1999) Vascular sources of phenylalanine, tyrosine, lysine, and methionine for casein synthesis in lactating goats. Journal of Dairy Science 82, 362–377. Boll, M., Markovich, D., Weber, W.-M., Korte, H., Daniel, H. and Murer, H. (1994) Expression cloning of a cDNA from rabbit small intestine related to proton-coupled transport of peptides, ␤-lactam antibiotics and ACE-inhibitors. European Journal of Physiology 429, 146–149. Boll, M., Herget, M., Wagener, M., Weber, W.-M., Markovich, D., Biber, J., Clauss, W., Murer, H. and Daniel, H. (1996) Expression cloning and functional characterization of the kidney cortex high-affinity proton-coupled transporter. Proceedings of the National Academy of Sciences USA 93, 284–289. Bowers, S.H. (1997) Characterization of glycyl-sarcosine uptake by ovine intestinal brush border membrane vesicles. MS Thesis, Virginia Polytechnic Institute and State University, Blacksburg, Virginia. Brandsch, M.C., Ganapathy, M.E., Chew, C.S., Ganapathy, V. and Leibach, F.H. (1997) Influence of proton and essential histidyl residues on the transport kinetics of the H+/peptide cotransport systems in intestine (PEPT1) and kidney (PEPT2). Biochimica et Biophysica Acta 1324, 251–262. Chen, H., Wong, E.A. and Webb, K.E., Jr (1999) Tissue distribution of a peptide transporter mRNA in sheep, dairy cows, pigs, and chickens. Journal of Animal Science 77, 1277–1283. Daniel, H. and Adibi, S.A. (1995) Selective effect of zinc on uphill transport of oligopeptides into kidney brush border membrane vesicles. FASEB Journal 9, 1112–1117. Daniel, H. and Herget, M. (1997) Cellular and molecular mechanisms of renal peptide transport. American Journal of Physiology 273, F1–F8. Daniel, H., Amasheh, S., Wenzel, U., Weber, W.-M. and Clauss, W. (1996) Transport of charged substrates by the intestinal peptide carrier PepT1 expressed in Xenopus laevis oocyte. FASEB Journal 10, A121 (Abstract). Fei, Y.-J., Kanal, Y., Nussberger, S., Ganapathy, V., Leibach, F.H., Romero, M.F., Singh, S.K., Boron, W.F. and Hediger, M.A. (1994) Expression cloning of a mammalian proton-coupled oligopeptide transporter. Nature 368, 563–566. Fei, Y.-J., Liu, W., Prasad, P.D., Kekuda, R., Oblak, T.G., Ganapathy, V. and Leibach, F.H. (1997) Identification of the histidyl residue obligatory for the catalytic activity of the human H+/peptide cotransporters PEPT1 and PEPT2. Biochemistry 36, 452–460. Freeman, T.C., Bentsen, B.S., Thwaites, D.T. and Simmons, N.L. (1995) H+/di-tripeptide transporter (PepT1) expression in the rabbit intestine. European Journal of Physiology 430, 394–400.

Tissue, Cellular and Molecular Aspects of Peptide Absorption and Utilization

129

Ganapathy, V. and Leibach, F.H. (1985) Is intestinal peptide transport energized by a proton gradient? American Journal of Physiology 249, G153–G160. Jayawardena, V. (2000) Ruminal and omasal absorption of ruminally derived amino acids and peptides in vitro. PhD Dissertation, Virginia Polytechnic Institute and State University, Blacksburg, Virginia. Koeln, L.L., Schlagheck, T.S. and Webb, K.E., Jr (1993) Amino acid flux across the gastrointestinal tract and liver of calves. Journal of Dairy Science 76, 2275–2286. Liang, R., Fei, Y.-J., Prasad, P.D., Ramamoorthy, S., Han, H., Yang-Feng, T.L., Hediger, M.A., Ganapathy, V. and Leibach, F.H. (1995) Human intestinal H+/peptide cotransporter cloning, functional expression, and chromosomal localization. Journal of Biological Chemistry 270, 6456–6463. Lister, N., Bailey, P.D., Collier, I.D., Boyd, C.A.R. and Bronk, J.R. (1997) The influence of luminal pH on transport of neutral and charged dipeptides by rat small intestine, in vitro. Biochimica et Biophysica Acta 1324, 245–250. Liu, W., Liang, R., Ramamoorthy, S., Fei, Y.-J., Ganapathy, M.E., Hediger, M.A., Ganapathy, V. and Leibach, F.H. (1995) Molecular cloning of PEPT 2, a new member of the H+/ peptide cotransporter family, from human kidney. Biochimica et Biophysica Acta 1235, 461–466. McCollum, M.Q. and Webb, K.E., Jr (1998) Glycyl-L-sarcosine absorption across ovine omasal epithelium during coincubation with other peptide substrates and volatile fatty acids. Journal of Animal Science 76, 2706–2711. Madara, J.L. and Pappenheimer, J.R. (1987) Structural basis for physiological regulation of paracellular pathways in intestinal epithelia. Journal of Membrane Biology 100, 149. Matthews, J.C. and Webb, K.E., Jr (1995) Absorption of L-carnosine, L-methionine, and Lmethionylglycine by isolated sheep ruminal and omasal epithelial tissue. Journal of Animal Science 73, 3464–3475. Matthews, J.C., Pan, Y.L., Wang, S., McCollum, M.Q. and Webb, K.E., Jr (1996a) Characterization of gastrointestinal amino acid and peptide transport proteins and the utilization of peptides as amino acid substrates by cultured cells (myogenic and mammary) and mammary tissue explants. In: Kornegay, E.T. (ed.) Nutrient Management of Food Animals to Enhance the Environment, CRC Press, Inc., Boca Raton, Florida. Matthews, J.C., Wong, E.A., Bender, P.K., Bloomquist, J.R. and Webb, K.E., Jr (1996b) Demonstration and characterization of dipeptide transport system activity in sheep omasal epithelium by expression of mRNA in Xenopus laevis oocytes. Journal of Animal Science 74, 1720–1727. Miyamoto, K.-I., Shiraga, T., Morita, K., Yamamoto, H., Haga, H., Taketani, Y., Tamai, I., Sai, Y., Tsuji, A. and Takeda, E. (1996) Sequence, tissue distribution and developmental changes in rat intestinal oligopeptide transporter. Biochimica et Biophysica Acta 1305, 34–38. Muller, U., Brandsch, M., Prasad, P.D., Fei, Y.-J., Ganapathy, V. and Leibach, F.H. (1996) Inhibition of the H+/peptide cotransporter in the human intestinal cell line Caco-2 by cyclic AMP. Biochemical and Biophysical Research Communications 218, 461–465. Ogihara, H., Saito, H., Shin, B.-C., Terada, T., Takenoshita, S., Nagamachi, Y., Inui, K.-I. and Takata, K. (1996) Immuno-localization of H+/peptide cotransporter in rat digestive tract. Biochemical and Biophysical Research Communications 220, 848–852. Pan, Y.X. (1999) Structural and functional characteristics of ovine PepT1. PhD Dissertation, Virginia Polytechnic Institute and State University, Blacksburg, Virginia. Pan, Y. and Webb, K.E., Jr (1998) Peptide-bound methionine as methionine sources for protein accretion and cell proliferation in primary cultures of ovine skeletal muscle. Journal of Nutrition 128, 251–256. Pan, Y., Bender, P.K., Akers, R.M. and Webb, K.E., Jr (1996) Methionine-containing peptides can be used as methionine sources for protein accretion in cultured C2C12 and MAC-T cells. Journal of Nutrition 126, 232–241.

130

K.E. Webb, Jr Pan, Y.-X., Wong, E.A., Bloomquist, J.R. and Webb, K.E., Jr (1997) Poly(A)+ RNA from sheep omasal epithelium induces expression of a peptide transport protein(s) in Xenopus laevis oocytes. Journal of Animal Science 75, 3323–3330. Pan, Y., Bender, P.K., Akers, R.M. and. Webb, K.E., Jr (1998) One or more serum factors promote peptide utilization in cultured animal cells. Journal of Nutrition 128, 744–750. Saito, H., Okuda, M., Terada, T., Sasaki, S. and Inui, K.-I. (1995) Cloning and characterization of a rat H+/peptide cotransporter mediating absorption of beta-lactam antibiotics in the intestine and kidney. Journal of Pharmacology and Experimental Therapeutics 275, 1631–1637. Saito, H., Terada, T., Okuda, M., Sasaki, S. and Inui, K.-I. (1996) Molecular cloning and tissue distribution of rat peptide transporter PEPT2. Biochimica et Biophysica Acta 1280, 173–177. Seal, C.J. and Parker, D.S. (1996.) Effect of intraruminal propionic acid infusion on metabolism of mesenteric- and portal-drained viscera in growing steers fed a forage diet: II. Ammonia, urea, amino acids, and peptides. Journal of Animal Science 74, 245–256. Shennan, D.B., Calvert, D.T., Backwell, F.R.C. and Boyd, C.A.R. (1998) Peptide aminonitrogen transport by the lactating rat mammary gland. Biochimica et Biophysica Acta 1373, 252–260. Steel, A., Nussberger, S., Romero, M.F., Boron, W.F., Boyd, C.A.R. and Hediger, M.A. (1997) Stoichiometry and pH dependence of the rabbit proton-dependent oligopeptide transporter PepT1. Journal of Physiology 498, 563–569. Wang, S. (1994) Peptides as amino acid sources for the synthesis of secreted proteins by mammary tissue explants and cultured mammary epithelial cells. PhD Dissertation, Virginia Polytechnic Institute and State University, Blacksburg, Virginia. Wang, S., Webb, K.E., Jr and Akers, M.R. (1996) Peptide-bound methionine can be a source of methionine for the synthesis of secreted proteins by mammary tissue explants from lactating mice. Journal of Nutrition 126, 1662–1672. Webb, K.E., Jr (1986) Amino acid and peptide absorption from the gastrointestinal tract. Federal Proceedings 45, 2268–2271. Webb, K.E., Jr (1990) Intestinal absorption of protein hydrolysis products: A review. Journal of Animal Science 68, 3011–3022. Webb, K.E., Jr and Bergman, E.N. (1991) Amino acid and peptide absorption and transport across the intestine. Proceedings of the 7th International Symposium on Ruminant Physiology, pp. 111–128. Webb, K.E., Jr and Matthews, J.M. (1994) Absorption of amino acids and peptides In: Asplund, J.M. (ed.) Amino Acid Nutrition of Ruminants. CRC Press, Boca Raton, Florida. Webb, K.E., Jr and Matthews, J.C. (1998) Peptide absorption and its significance in ruminant protein nutrition. In: Grimble, G. and Backwell, C. (eds) Peptides in Mammalian Protein Metabolism: Tissue Utilization and Clinical Targeting. Portland Press Ltd, London, pp. 1–10. Webb, K.E., Jr, Matthews, J.C. and DiRienzo, D.B. (1992) Peptide absorption: A review of current concepts and future perspectives. Journal of Animal Science 70, 3248–3257. Webb, K.E., Jr, DiRienzo, D.B. and Matthews, J.C. (1993) Recent developments in gastrointestinal absorption and tissue utilization of peptides – A review. Journal of Dairy Science 76, 351–361. Wenzel, U., Gebert, I. and Daniel, H. (1996) Characteristics of uptake of neutral and anionic cephalosporin antibiotics by the cloned intestinal peptide transporter PePT1. FASEB Journal 10, A283 (abstract). Wolffram, S., Grenacher, B. and Scharrer, E. (1998) H+-coupled uphill transport of the dipeptide glycylsarcosine by bovine intestinal brush-border membrane visicles. Journal of Dairy Science 81, 2595–2603.

8

Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver C.J. SEAL AND D.S. PARKER Department of Biological and Nutritional Sciences, Faculty of Agriculture and Biological Science, University of Newcastle upon Tyne, Newcastle upon Tyne, UK

Introduction Tissues of the splanchnic bed include the gastrointestinal tract (GIT), liver, spleen, pancreas and mesenteric fat depots. Cumulatively these organs, together with the associated connective tissue and blood vessels, contribute approximately 15–20% of total body mass. During pregnancy and lactation, hypertrophy of the GIT can increase this still further. Collectively the splanchnic bed plays a central role in moderating nutrient supply to peripheral tissues for maintenance and productive processes such as muscle deposition, wool growth and milk production. Liver output is affected by the pattern and quantity of nutrients delivered to the portal vein from the GIT and this process is discussed separately in Chapter 9. Intestinal tissues maintain a high rate of metabolic activity which is sustained by both the luminal supply of nutrients available during the process of absorption and also from metabolites derived from the arterial blood supply. This chapter will review the effects of GIT metabolism on the supply of energy-yielding substrates to the liver, and will examine the effects of changing diet on the supply of nutrients such as short-chain volatile fatty acids (SCVFA), glucose and amino acids to the liver.

Tissues of the splanchnic bed Anatomy and vasculature of the gastrointestinal tract and liver The common blood vessel supplying the liver is the portal vein which divides into the lobes of the liver at the porta hepatis plexus (Seal and Reynolds, 1993). Blood arriving at the liver in the portal vein represents the cumulative drainage of venous blood from the entire GIT and its associated organs including the pancreas, spleen and mesenteric fat. Blood sampled at the porta hepatis and compared with arterial blood, therefore, represents the cumulative effects of metabolism across a range of diverse metabolically © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

131

132

C.J. Seal and D.S. Parker

active tissues. For the GIT, the additional effects of compartmentalization of the digestive process along the tract cannot be identified with reference to portal blood alone. Technically, the biggest advance in understanding splanchnic metabolism has been the development of in vivo, multi-catheterization procedures to sample portal blood and blood draining other areas of the GIT. Elegantly described by Huntington et al. (1989) for cattle, this has provided the benchmark for much of the work in this field over the past decade, following the early work of Bergman’s group with sheep (Katz and Bergman, 1969). Anatomically, however, there are significant differences between cattle and sheep, which has made direct comparison of catheterization techniques difficult. The common portal vein of cattle, for example, is much shorter than in sheep making the implantation of transonic flow probes complex (see later). The proximity of the gastro-splenic vein also makes location of the catheter tip difficult (Huntington, 1982). Additionally, the venous drainage of sheep is not associated with discrete gut components as observed with the closed loop draining into a ‘mesenteric’ vein in cattle (Seal and Reynolds, 1993). In an attempt to resolve this problem in sheep, Neutze et al. (1994) have described a complex procedure in which they have measured ␣-amino–N uptake in the caudal mesenteric vein upstream and downstream of the ileocaecocolic vein junction. The surgery involved in this procedure is technically demanding, requiring the precise location of catheters in small blood vessels. Their results, however, were very comparable with similar data from mesenteric vein samples obtained from cattle, suggesting that this may be a reasonable approach for future studies.

Measurement of blood flow Measuring blood flow through catheterized vessels is a key parameter for the determination of net nutrient fluxes across the splanchnic bed. Errors in the determination of blood flow, and variability in its measurement, are critical factors in all studies of this type. For many metabolites, small venous–arterial (V–A) differences are amplified by large blood flows to give widely variable (and perhaps unreliable) flux rates. This problem is further exacerbated by the increasing use of more ‘sensitive’ analytical procedures coupled with labelled metabolites leading to V–A differences that are at the limit of analytical precision. The majority of blood flow data reported in the literature have been obtained using dye-dilution procedures, the most common being dilution of p-amino hippuric acid (PAH). The method was initially described in sheep by Katz and Bergman (1969) and involves the use of a primed, continuous infusion of PAH upstream of the sampling catheter through a second infusion catheter, usually implanted in a small mesenteric vein. Alternative dyes used in other animal species and man include indocyanine green (ICG; Shoemaker, 1964; Wangsness and McGilliard, 1972; Ferrannini et al., 1988) and antipyrine (Marconi et al., 1989). Blood flow is calculated by determining the difference in indicator dye concentration between the target vessel and the peripheral circulation relative to the dye infusion rate. Corrections for water absorption/resorption across the gut tissues can be made using differences in haematocrit between splanchnic blood and peripheral blood. In our experience, however, this correction was small and inconsistent and for the majority of our studies was not significant. Accurate deter-

Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver

133

mination of blood flow, regardless of the dye used, is dependent on complete mixing of the dye with the blood before it reaches the tip of the sampling catheter. This is usually achieved by implanting the infusion catheter several centimetres distal to the sampling tip. However, ‘streaming’ of blood through vessels may still be a major factor in causing variability in this data and can be reduced by placing the catheter tip in the lateral lobe of the liver beyond the turbulent flow of the porta hepatis (Seal and Reynolds, 1993). Direct measurement of portal and mesenteric vein blood flow is possible through the use of transonic flow probes. This procedure has been in use for more than 30 years (Carr and Jacobson, 1968), but has been limited by technical difficulties during surgical implantation and in calibration. This is a particular problem for large cattle, where the size and accessibility of the portal vein severely restricts the positioning of the probe (Huntington et al., 1990). The size of the probe necessary to encase the portal vein in these animals may also influence the transmission of the Doppler signal across the vein, giving false measurements of blood flow. The procedure has been used with greater success in sheep (Neutze et al., 1994; Remond et al., 1998) and in small growing steers (Kim et al., 1998). In some of these studies it has been reported that blood flow determined by ultrasonic transit-time flow meters underestimates blood flow compared with indicator dilution techniques (Huntington et al., 1990; Kristensen et al., 1996; Remond et al., 1998), but this observation is not consistent, and may be improved with the use of newer types of probe (Remond et al., 1998). This underestimate is presumed to be due to turbulence in the blood flow through the probe. The principal advantage of the transonic probes is the ability to measure blood flow continuously over long time periods using automated data-handling systems. This reduces the frequency with which blood samples are taken, and gives an ‘integrated’ blood flow profile which can be used to describe and model diurnal variations. Measurements of this type in steers show that blood flow in twice-daily-fed animals varies in a sinusoidal manner (Fig. 8.1; Kim et al., 1998) which is similar to data obtained using frequent blood sampling and complex non-parametric curve-fitting methods (Whitt et al., 1996). Future developments in transonic probe technology will further increase the value of this methodology.

Use of labelled substrates Measurements of V–A concentration differences and blood flow across tissues of the GIT and liver have been invaluable in providing information linking the digestion and absorption of feed in the gut lumen and the subsequent appearance of nutrients in the venous drainage. However, flux rates thus determined represent the net effect of transfers across the apical and basolateral surfaces of the intestinal cells and metabolism within the cell. Positive V–A differences result from a net flux of a nutrient from lumen to blood that exceeds the sum of extraction from blood and utilization within the intestinal tissues. Conversely, a negative V–A indicates that extraction from arterial blood and utilization within the intestinal tissues is greater than the flux of the nutrient from the gut lumen. For some nutrients, for example glucose, V–A may fluctuate from positive to negative, depending on the diet (see, for example, Seal et al., 1992, 1994; Seal and Parker, 1994; Balcells et al., 1995; Piccioli Cappelli et al., 1997). For others,

134

C.J. Seal and D.S. Parker

Fig. 8.1. A typical circadian pattern for portal vein blood flow in a 150 kg Holstein–Friesian steer (feed offered at 60 and 780 min). Reproduced with permission from Kim et al. (1998).

for example acetate, net fluxes are positive across ruminal tissues, but may be negative across mesenteric-drained tissues (Seal et al., 1992; Seal and Parker, 1994). Hind-gut fermentation may also result in positive fluxes of SCVFA into the caecal vein (DeGregorio et al., 1984). Isotopic labelling of nutrients in combination with classic V–A techniques provides the opportunity to estimate true rates of absorption from the gut and can be used to determine rates of utilization of nutrients within gut tissues as well as sequestration of nutrients from the blood supplying the gut. Combinations of luminal and vascular infusions of labelled substrates can be used to quantify whole-body irreversible loss rates for individual nutrients, inter-conversion of metabolites and metabolism across tissues in single experiments, thus maximizing the data produced from a multi-faceted approach. If these studies are further combined with surgical cannulation of the gut tissues, direct measurements of intestinal disappearance of nutrients can be compared with simultaneous measurements of nutrient flux into the venous drainage. Improvements in analytical procedures and developments in stable isotope technologies will undoubtedly further increase the potential for this type of study.

Energy metabolism in the gastrointestinal tract Estimates of the proportion of total oxygen consumption attributed to the tissues of the whole splanchnic bed range from 35 to 60% (Seal and Reynolds, 1993) and

Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver

135

approximately 20% for the GIT alone (Cant et al., 1996). These tissues account for about 10–13% of total body mass, and thus, on a unit mass basis, their metabolic activity is considerable. Quantitatively, the major components of this energy expenditure are Na+, K+-ATPase-linked ion transport systems (approximately 30–60% of the expenditure) and protein turnover (20–23% for protein synthesis and 4% for protein degradation; McBride and Kelly, 1990). Factors which influence energy expenditure include meal consumption (Christopherson and Brockman, 1989; Kelly et al., 1989) and energy intake expressed as either metabolizable energy (see Seal and Reynolds, 1993) or digestible energy (Goetsch and Patil, 1997; Goetsch, 1998). Physiological state also has a major impact on intestinal energy expenditure. For example, pregnancy and lactation are associated with increased intestinal tissue mass (Fell et al., 1972; Hammond and Diamond, 1994). This latter response, observed in both ruminant and non-ruminant species, results in increased mucosal mass and concomitant increases in protein synthetic rates in GIT tissues. This increased intestinal transport capacity also results in the synchronous upregulation of all intestinal transporters (Bird et al., 1996). The relationships between digestible energy intake, fibre digestion and urea-N flux have been described in some detail by Goetsch (1998) in a series of regression equations based on 11 separate experiments with sheep. These data show that, in forage-fed animals, there is a complex interaction between portal-drained viscera energy use and digestible energy intake and neutral detergent fibre (NDF) digestion; these factors and hepatic urea net flux also impact on total splanchnic energy use and increase with increased forage consumed. The combined effects of these variables, especially the increased energy required in hepatic ureagenesis to dispose of ruminal ammonia (Lobley et al., 1996b), may account for the observed inefficiency of growth observed in animals fed forage-based diets. The processes involved in changing GIT energy expenditure, and the consequences on animal performance, have been the focus of several recent reviews (Lobley, 1994; Cant et al., 1996; Goetsch, 1998).

Short-chain volatile fatty acid metabolism Quantitatively, the major source of absorbed energy in ruminants is in the form of SCVFAs (acetate, propionate and butyrate). The net flux of these SCVFAs across the portal-drained viscera is highly correlated with both the composition of the diet and the level of feed intake (Huntington, 1983; Huntington and Prior, 1983). For any single diet at different levels of intake, the proportions of SCVFA appearing in portal blood remains constant, reflecting ruminal production rates for the individual acids. Changing ruminal fermentation pattern, by altering the forage to concentrate ratio of the diet, results in a concomitant change in the pattern of SCVFA in ruminal fluid and portal blood, but the extent of metabolism of the individual acids remains constant (Seal and Reynolds, 1993). Studies in which direct comparison between measured production rates of SCVFAs in the rumen and net flux into the portal vein have been made are not common. Those that are available, however, confirm in vitro studies which show that there is extensive metabolism of all the acids during absorption across the gut wall.

136

C.J. Seal and D.S. Parker

Absorption of SCVFA SCVFA may be absorbed across the rumen epithelium in both undissociated and dissociated forms. Undissociated SCVFAs are lipid soluble and cross the epithelial membrane more easily than the dissociated form; absorption rates of the undissociated acids are thus affected by the pH of ruminal fluid although decreasing pH only results in a small increase in SCVFA clearance (Dijkstra et al., 1993). In contrast to non-ruminant colonic and caecal tissue (Holtug et al., 1992), Na+/H+ exchange is not involved in SCVFA absorption (López et al., 1996). Dissociated SCVFAs are absorbed via anion exchange with bicarbonate, in a mechanism which appears similar for non-ruminant large intestinal and ruminal tissues and involves an additional electroneutral anion exchange between bicarbonate and chloride.

SCVFA metabolism by gut tissues Within ruminal tissues, acetate is primarily metabolized for energy; propionate contributes to energy supply, lactate and CO2; and butyrate is extensively metabolized to 3-hydroxy butyrate, CO2 and energy (Bergman, 1990; Britton and Krehbiel, 1993). The initial stage of SCVFA metabolism involves the formation of activated CoA-esters of the acids, and the extent of their metabolism depends on the activities of two acylCoA-synthetases present in the epithelial cells: one which is non-specific for all three major SCVFAs and one which is specific for butyrate. This specificity ensures that butyrate is preferentially metabolized in the rumen epithelium. Although the extent of acetate metabolism is lower than that for the other SCVFAs, quantitatively it may account for the majority of energy expended by ruminal tissues due to the high ruminal production rates of acetate compared with propionate and butyrate (Britton and Krehbiel, 1993). Metabolism of SCVFA during absorption increases with increasing chain length with between 40 and 69% of acetate, 30 and 78% of propionate and 8 and 66% of butyrate ruminal production appearing in portal blood (Table 8.1). Care should be taken in interpreting measured rates of SCVFA appearance into portal blood which may underestimate true rates of absorption from the rumen. For example, acetate is utilized by the mesenteric drained viscera (Bergman and Wolff, 1971; Seal et al., 1992; Seal and Parker, 1994; C.J. Seal, D.S. Parker, J.C. MacRae and G.E. Lobley, unpublished data) as shown by negative V–A in the mesenteric vein which must be balanced, and then exceeded by acetate absorbed from the rumen. In addition for some diets, microbial fermentation of residual carbohydrate in the hind-gut may contribute SCVFA to portal blood (DeGregorio et al., 1984; Owens et al., 1986), which cannot be distinguished from SCVFA derived from the rumen. In our studies in steers with mesenteric vein catheters, we have been unable to detect positive V–A for propionate across post-ruminal tissues. This suggests that although approximately 30% of propionate produced in the rumen may escape into the abomasum and omasum (Peters et al., 1990), this is completely metabolized within the tissues of these gut compartments. Further studies using labelled SCVFA are needed to elucidate the processes of absorption and metabolism of these nutrients during passage through, and absorption from, different compartments of the GIT.

Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver

137

Table 8.1. Net portal flux of acetate, propionate and butyrate in sheep and cattle expressed as a proportion of measured ruminal production rate. (Updated from Seal and Reynolds, 1993.) Proportion of rumen production rate appearing in portal vein Animal

Diet

Lambs

Intragastric infusion

Lambs

Intragastric infusion

Sheep

Intragastric infusion

Sheep Sheep

Forage High fibre Low fibre Concentrate Concentrate Forage Concentrate Forage + 0.5 mol propionate day⫺1 + 1 mol propionate day⫺1 Concentrate + butyrate

Cattle Cattle Steers Steers

Steers

Acetate Propionate Butyrate 0.45 0.40 0.53 0.53 0.60 0.44

0.69 0.62 0.57 0.69 0.43 0.48 0.52 0.96 0.74 0.76

0.77 0.52 0.78 0.66 0.66 0.67 0.57 0.59 0.49 0.51 0.58 0.49 0.40 0.30 0.30 0.52 0.56 0.58

0.25 0.14 0.15 0.28 0.28 0.26

Referencea a

b

c 0.08

0.08 0.66

d e e f g h i

0.47 0.31 0.34 0.32 0.30 0.29

j

aData

from: a, Gross et al., 1990b; b, Gross et al., 1990a; c, Weekes and Webster, 1975; d, Bergman, 1990; e, Calculated from Linington et al., 1998a, and Linington et al., 1998b; f, Harmon et al., 1988; g, Calculated from Huntington and Reynolds, 1983 and Sharp et al., 1982); h, Seal et al., 1992; i, Seal and Parker, 1994; j, Krehbiel et al., 1992.

Glucose metabolism Glucose metabolism by gut tissues Glucose metabolism in ruminant species is dominated by the requirement for suitable precursors for gluconeogenesis, reflecting the lack of glucose absorbed from the digestive tract of forage-fed animals. Glucose requirements for tissue metabolism are similar to those of other species (Weekes, 1991), although glucose sparing from cellular activity such as fatty acid synthesis limits carbon flux through key metabolic pathways. Experiments which measure overall glucose utilization by the animal and the contribution of different tissues to that flux, show that the gut plays a major role in these processes. Glucose has been shown to be an energy substrate for the intestinal mucosa

138

C.J. Seal and D.S. Parker

(Stangassinger and Giesecke, 1986; Britton and Krehbiel, 1993; Okine et al., 1993) and calculations of net glucose utilization by ruminants fed a range of diets (Parker, 1990) demonstrated that, in most dietary situations, the gut is a net consumer of glucose (negative V–A difference across the tissue). In studies in pregnant and lactating sheep (Van der Walt et al., 1983; Perry et al., 1994), glucose utilization by both the mesenteric-drained viscera (MDV) and portal-drained viscera (PDV) increased in animals post partum, with a concomitant rise in lactate production. In studies with sheep and steers where glucose irreversible loss (GIL) has been manipulated by either increasing glucose or propionate supply, a similar relationship between portal glucose utilization and GIL has been shown (Seal and Parker, 1994; Balcells et al., 1995; Piccioli Cappelli et al., 1997); the results of these studies are summarized in Fig. 8.2. These data clearly show that the use of glucose by gut tissues increases in line with whole body turnover rate. Further work in our laboratory has also shown that glucose utilization by gut tissues is responsive to changes in both vascular and luminal glucose supply (Piccioli Cappelli et al., 1997), underlining the ability of the tissue to capitalize on both sources of the nutrient.

Fig. 8.2. Relationship between whole body glucose irreversible loss rate (mmol min⫺1) and portal glucose utilization rate (mmol min⫺1) in sheep and steers (data from Balcells et al., 1995, and C.J. Seal, D.S. Parker, J.C. MacRae and G.E. Lobley, unpublished data; y = 0.25x + 0.007, r = 0.642).

Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver

139

Starch digestion, glucose release and absorption In most dietary situations, the availability of glucose in the gut lumen is low (Armstrong and Smithard, 1979) and gut tissue requirements are met from the arterial supply. The inclusion of significant amounts of raw materials such as maize, either as ensiled forage or as grain, in ruminant diets, however, has been shown to result in the flow of undegraded starch to the small intestine. These observations have provoked considerable discussion as to the ability of the ruminant gut to digest starch and absorb the glucose monomers (Nocek and Tamminga, 1991). A review of digestibility studies investigating the extent and site of digestion of starch within the ruminant gut (Huntington, 1997) gives total tract values of between 87 and 99% with, on average, 5–20% of starch consumed digested post-ruminally (Fig. 8.3). A number of experiments in this area have provided evidence that there may be a limitation in the ability of the ruminant small intestine to digest high levels of the polymer. It has been suggested that this restriction is linked to low levels of secretion of pancreatic amylase (Hill et al., 1991; Kreikemeier et al., 1991; Harmon, 1992). This hypothesis has been analysed in some detail by Huntington (1997) in a series of simulations of starch flow and digestion incorporating digesta flow measurements with net glucose absorption data and values for glucose transport kinetics across the small intestine. These calculations suggest that gut glucose transporter capacity is unlikely to limit glucose uptake

Fig. 8.3. Relationship between starch intake (g day⫺1) and digestibility (expressed as a percentage of intake) in the rumen (□; y = ⫺4.24x + 95.83, r = 0.793), postrumen (䢇; y = 3.61x + 1.35, r = 0.859) and total tract (䊱; y = ⫺0.64x + 97.66, r = 0.404) of beef and dairy cattle. Data from Huntington (1997).

140

C.J. Seal and D.S. Parker

and that in many cases the capacity of the ruminant small intestine to absorb glucose is in excess of supply. These data also suggest that any reduction in starch digestibility is a result of insufficient amylase activity and that stimulation of amylase synthesis/ release would overcome this limitation. Recent experiments investigating the expression of SGLT1 Na+/glucose co-transporter gene in gut tissues from lactating cows (Zhao et al., 1998) confirm that the epithelial cells throughout the digestive tract, including the rumen and omasum, have the capacity to actively transport glucose. The kinetics of the transport process show that the system has a high affinity for D-glucose with a Km of about 0.1 mM confirming earlier studies by Bauer et al. (1995). In these experiments, ruminants unadapted to digesting starch were shown to have an ability to transport glucose which was blocked by the presence of phlorizin. In contrast to earlier work with sheep (Shirazi-Beechey et al., 1991), these later experiments demonstrated that both cattle and sheep maintained a capacity for glucose uptake by the small intestine in the absence of a flow of dietary starch to the small intestine, an observation confirmed in work in our own laboratory (Balcells et al., 1995; Piccioli Cappelli et al., 1997). The relationship between transporter number at the apical and basolateral membranes and gut capacity to transport glucose has been reviewed (Ferraris et al., 1992; Croom et al., 1998) and endocrine factors involved in transport regulation identified (Bird et al., 1996). Almost all these studies have been carried out on monogastric species, apart from one in which sheep treated with bST showed up-regulation of Na+-dependent glucose transport in the duodenum (Bird et al., 1994). This effect was not apparent in jejunal or ileal tissue indicating a site-specific response, possibly promoted by release of factor(s) such as epidermal growth factor.

Effects of glucose availability on nutrient flux The site and extent of starch digestion in the ruminant gut has an influence on the absorbed end products of digestion and also the flow of other nutrients within the tract. Studies such as that by Reynolds et al. (1991) provide clear evidence of the impact that forage versus concentrate diets have upon the pattern of nutrient uptake and subsequent animal performance. Greater energy retention in concentrate-fed heifers was linked to changes in the metabolism of visceral tissues, including an increase in net portal-drained visceral glucose flux as a result of increased starch flow to the small intestine. A similar effect was reported by Goetsch and Ferrell (1995), who fed different forage sources with increasing levels of maize and monitored nutrient flux across the gut. Increasing dietary maize level had no effect on PDV glucose flux. However, for two of the forage sources, propionate release increased linearly with maize inclusion, whereas for the third, a linear response in PDV release of lactate was reported. These data underline the impact that diet composition has upon the site of digestion and the subsequent availability of substrates for gut tissue metabolism. These ideas have been extended in a number of interesting studies in which combinations of glucose and protein infusions into the rumen and small intestine have been undertaken to attempt to define the interactions between site and nature of nutrients on animal responses. In the work reported by Taniguchi et al. (1995), starch and casein were infused into either the rumen or the abomasum and net absorption of the end products

Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver

141

of digestion monitored across the gut. Energy release by the PDV per unit of digestible energy intake was the same, whichever site was used for starch infusion. Net release of glucose by the PDV, however, was higher during simultaneous infusion of starch and casein into the abomasum, when compared with abomasal starch infusion and intraruminal casein infusion. These data support an earlier hypothesis that increased protein flow to the duodenum stimulated pancreatic enzyme release, resulting in increased digestibility of starch escaping fermentation in the rumen. Analysis by Huntington (1997) on the impact of maize inclusion in the diet of feedlot steers and high-yielding dairy cows on glucose kinetics and the use of glucose by the gut tissues identifies the contribution of post-ruminal starch digestion to splanchnic tissue metabolism. Although this is significant in both production situations, it is apparent that for the growing steer increased fermentation of the available starch in the rumen, resulting in energy substrate release and an increase in microbial protein flow might improve the overall efficiency of nutrient use for growth.

Amino acid metabolism Amino acid requirements of gut tissues Although the tissues of the gut represent only a small proportion of body protein mass (0.05 in cattle), they contribute a disproportionate amount to whole body protein synthesis (0.32–0.45, Lobley et al., 1980). In addition, when compared to muscle, gut tissues are characterized by having a high rate of protein turnover and a low level of protein accretion, reflecting the ability to adapt rapidly to changing nutritional situations so as to maintain a supply of nutrients to the whole animal. The impact of changes in gut function during the transition from milk-fed to weaning in lambs, for example, has been shown to result in enhanced rates of protein synthesis throughout the tract (Attaix et al., 1992), reflecting the increased intake of a more fibrous diet of lower digestibility. This adaptive process has been suggested as one of the reasons why ruminant animals show a low apparent efficiency of utilization of absorbed amino acids when compared to monogastric species. This hypothesis was supported by the pioneering work on amino acid uptake in sheep (Tagari and Bergman, 1978) in which apparent disappearance of amino acids from the gut was compared to the net absorption measured in the portal vein. These studies suggested that, for a number of the essential amino acids, less than 50% of the measured loss from the gut could be accounted for in terms of net absorption into the portal vein. Since that time, there has been a continued interest in the extent to which amino acid metabolism by the tissues of the gut impacts upon whole body protein accretion and the factors which influence this effect. Our understanding of the processes involved has been considerably enhanced by the development of infusion and tracer techniques using mixed amino acid substrates such as 13C algal protein hydrolysates (Lobley et al., 1996a). This technique has been applied in an experiment to investigate the relationship between luminal and arterial supply of amino acids to the gut tissues (MacRae et al., 1997) by sequential infusion of labelled amino acids into either the lumen of the gut or intravascularly. These data have shown that, in contrast to our previous understanding, luminal amino acids are efficiently transported into portal blood and that the majority (80%) of the amino acids

142

C.J. Seal and D.S. Parker

sequestered in the gut tissue arise from the arterial supply to the gut. These data are shown in Fig. 8.4, which includes data for the proportion of whole body flux of selected amino acids which is utilized by the gut. These are important observations because they indicate that, far from capitalizing on the availability of dietary amino acids, the GIT is effectively competing with the other tissues of the body for arterially supplied amino acids, similar to the situation with glucose (see above). Recent work from Lobley et al. (1998) in which increased levels of a mixture of amino acids were infused into the mesenteric vein tends to support this hypothesis. During these infusions, net appearance of amino acids across the PDV was below the theoretical level, suggesting that either there was an inhibition of amino acid uptake from the gut or there was increased uptake of amino acids from the systemic circulation. Sequestration of amino acids in gut proteins represents a significant element in overall amino acid flux, whether it is derived from arterial or luminal supply. The extent to which this process can be manipulated in ruminant species is unclear. Studies in which steers were treated with recombinant somatotrophin during infusion of casein into the abomasum (Bruckental et al., 1997) demonstrate that hormone treatment can alter the partition of amino acids into gut tissues resulting, in this case, in reduced PDV flux. Similarly, Yu et

60 (0.81) (0.85)

50

(0.84)

(0.83) mmol day⫺1

40 (0.78) 30

(0.48)

20

(0.77)

10

0 Leu

Val

Lys

Thr

Iso

Phe

His

0.43

0.62

0.56

0.50

0.52

0.51

0.32

Proportion of whole body flux

Fig. 8.4. Rates of amino acid sequestration (mmol day⫺1) in gut tissues from arterial (open portion of bars) and luminal (hatched portion of bars) precursors in sheep fed 800 g day⫺1 lucerne pellets (reproduced with permission from MacRae et al., 1997).

Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver

143

al. (1998) showed that subclinical infection of Trichostongylus colubriformis in sheep increased leucine sequestration in gut tissues and reduced net release of absorbed leucine to the peripheral tissues; gut protein accretion in response to the parasite infection limiting synthetic processes elsewhere in the body. The extent to which the results from the study of MacRae et al. (1997) can be applied across a range of physiological and nutritional states has yet to be assessed. The use of a similar stable isotope infusion technique in baby pigs has provided somewhat conflicting data showing extensive oxidation of dietary amino acids during first pass through the gut tissues (Reeds, 1998). These latter observations may reflect differences between ruminant and non-ruminant species or changes associated with gut development. In this respect, the earlier work of Alpers (1972) in which 14C-leucine was infused either intraluminally or intravenously into the rat allowed the pattern of incorporation into enterocyte cells along the villus to be determined. The results showed that cells in the crypts and the base of the villus incorporated leucine from the blood whereas those at the tip of the villus and in the upper section utilized the labelled amino acid provided in the gut lumen. It is not known how such a differential use of amino acids might be influenced by situations in which enterocyte migration rate is altered during, for example, administration of gut active growth promoters (Parker, 1990).

Amino acid absorption and metabolism Amino acid use by the tissues of the ruminant gut has been studied using a number of models in which the pattern of nutrients has been manipulated or the demands of the tissue altered. In many of these studies, ruminant tissue oxidation of amino acids contributes only a limited extent to amino acid use, and the metabolism of glutamine, for example, may be different from that observed in non-ruminant species. Recent data from fasted sheep (Gate et al., 1999) showed that conversion of this amino acid to ammonia was much lower in sheep than in rats (Windmeuller and Spaeth, 1978) and that the fraction of glutamine extracted by the digestive tract to be catabolized is also lower in ruminants. Similar observations have been made for the branch-chain amino acids (Pell et al., 1986). Leucine utilization by gut tissues was shown to be directed primarily towards protein synthesis and these data are similar to our own obtained in sheep in which glucose flux, manipulated by intravenous or intraduodenal infusion of glucose, had no effect on overall leucine oxidation by gut tissues (Piccioli Cappelli et al., 1997). In earlier experiments in our laboratory, manipulation of glucose availability through intravenous infusion (Balcells et al., 1995) was shown to decrease net absorption of amino acids across the PDV in sheep, whereas propionate infusion into the rumen of steers resulted in an increase in amino acid flux across the MDV (Seal and Parker, 1996). These differences could reflect the impact that changes in the pattern of nutrient supply may have on metabolism by different parts of the gut, as determined by measurements made at different sites, i.e. PDV versus MDV (Reynolds and Huntington, 1988; Reynolds et al., 1994). The application of new isotope infusion techniques, and the recognition that the amino acid ‘nutrition’ of the gut represents flux from both arterial and luminal supplies, has advanced our appreciation of the impact that this active tissue has upon partition of amino acids within the whole animal. Quantification of these processes under different physiological and nutritional conditions

144

C.J. Seal and D.S. Parker

will provide a framework for improving the efficiency of protein utilization by ruminant species.

References Alpers, D.H. (1972) Protein synthesis in intestinal mucosa: the effect of route of administration of precursor amino acids. Journal of Clinical Investigation 51, 167–173. Armstrong, D.G. and Smithard, R.R. (1979) The fate of carbohydrates in the small and large intestines of the ruminant. Proceedings of the Nutrition Society 38, 283–294. Attaix, D., Aurousseau, E., Rosolowska-Huszcz, D., Bayle, G. and Arnal, M. (1992) In vivo longitudinal variations in protein synthesis in developing ovine intestines. American Journal of Physiology 263, R1318–R1323. Balcells, J., Seal, C.J. and Parker, D.S. (1995) Effect of intravenous glucose infusion on portaldrained visceral metabolism in sheep fed a cereal/straw-based diet. Journal of Animal Science 73, 2146–2155. Bauer, M.L., Harmon, D.L., McLeod, K.R. and Huntington, G.B. (1995) Adaptation to small intestinal starch assimilation and glucose transport in ruminants. Journal of Animal Science 73, 1828–1838. Bergman, E.N. (1990) Energy contributions of volatile fatty acids from the gastrointestinal tract in various species. Physiological Reviews 70, 567–590. Bergman, E.N. and Wolff, J.E. (1971) Metabolism of volatile fatty acids by liver and portaldrained viscera in sheep. American Journal of Physiology 221, 586–592. Bird, A.R., Croom, W.J., Fan, Y.K., Daniel, L.R., McBride, B.W. and Taylor, I.L. (1994) The effects of recombinant bovine somatotrophin on intestinal absorption in sheep. Journal of Animal Science 72 (Suppl 1), 70A. Bird, A.R., Croom, W.J., Fan, Y.K., Black, B.L., McBride, B.W. and Taylor, I.L. (1996) Peptide regulation of intestinal glucose absorption. Journal of Animal Science 74, 2523–2540. Britton, R. and Krehbiel, C. (1993) Nutrient metabolism by gut tissues. Journal of Dairy Science 76, 2125–2131. Bruckental, I., Huntington, G.B., Baer, C.K. and Erdman, R.A. (1997) The effect of abomasal infusion of casein and recombinant somatotropin hormone injection on nitrogen balance and amino acid fluxes in portal-drained viscera and net hepatic and total splanchnic blood in Holstein steers. Journal of Animal Science 75, 1119–1129. Cant, J.P., McBride, B.W. and Croom, W.J. (1996) The regulation of intestinal metabolism and its impact on whole animal energetics. Journal of Animal Science 74, 2541–2553. Carr, S.B. and Jacobson, D.R. (1968) Method for measurement of gastrointestinal absorption in normal animals, combining portal-carotid differences and telemetered portal blood flow by Doppler shift. Journal of Dairy Science 51, 721–729. Christopherson, R.J. and Brockman, R.P. (1989) Effect of feeding and a ␤-adrenergic blocking agent on oxygen consumption by the portal-drained viscera, liver and hind quarters of the sheep. In: Van der Honing, Y. and Close, W.H. (eds) Energy Metabolism of Farm Animals (EAAP Publication No 43). Wageningen, Netherlands: Centre for Agricultural Publishing and Documentation (Pudoc), pp. 147–150. Croom, W.J., McBride, B., Bird, A.R., Fan, Y.-K., Odle, J., Froetschel, M. and Taylor, I.L. (1998) Regulation of intestinal glucose absorption: a new issue in animal science. Canadian Journal of Animal Science 78, 1–13. DeGregorio, R.M., Tucker, R.E., Mitchell, G.E. and Gill, W.W. (1984) Acetate and propionate production in the caecum and proximal colon of lambs. Journal of Animal Science 58, 203–207. Dijkstra, J., Boer, H., Van Bruchem, J., Bruining, M. and Taminga, S. (1993) Absorption of

Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver

145

volatile fatty acids from the rumen of lactating dairy cows as influenced by volatile fatty acid concentration, pH and rumen liquid volume. British Journal of Nutrition 69, 385–396. Fell, B.F., Campbell, R.M., Mackie, W.S. and Weekes, T.E.C. (1972) Changes associated with pregnancy and lactation in some extra-reproductive organs in the ewe. Journal of Agricultural Science, Cambridge 79, 397–407. Ferrannini, E., DeFronzo, R.A., Gusberg, R., Tepler, J., Jacob, R., Aaron, M., Smith, D. and Barrett, E.J. (1988) Splanchnic amino acid and glucose metabolism during amino acid infusion in dogs. Diabetes 37, 237–245. Ferraris, R.P., Villenas, S.A. and Diamond, J. (1992) Regulation of brush-border enzyme activities and enterocyte migration rates in mouse small intestine. American Journal of Physiology 262, G1047–G1059. Gate, J.J., Parker, D.S. and Lobley, G.E. (1999) The metabolic fate of the amido-N group of glutamine in the tissues of the gastrointestinal tract in 24 h fasted sheep. British Journal of Nutrition 81, 297–306. Goetsch, A.L. (1998) Splanchnic tissue energy use in ruminants that consume forage-based diets ad libitum. Journal of Animal Science 76, 2737–2746. Goetsch, A.L. and Ferrell, C.L. (1995) Effects of dietary level of maize on net flux of oxygen and nutrients across splanchnic tissues in sheep consuming ad libitum different forages. Animal Science 61, 43–55. Goetsch, A.L. and Patil, A.R. (1997) Relationships among splanchnic tissue energy consumption and net flux of nutrients, feed intake and digestibility in wethers consuming forage-based diets ad libitum. Journal of Applied Animal Research 11, 1–18. Gross, K.L., Harmon, D.L. and Avery, T.B. (1990a) Portal-drained visceral flux of nutrients in lambs fed alfalfa or maintained by intragastric infusion. Journal of Animal Science 68, 214–221. Gross, K.L., Harmon, D.L., Minton, J.E. and Avery, T.B. (1990b) Effects of isoenergetic infusions of propionate and glucose on portal-drained visceral nutrient flux and concentrations of hormones in lambs maintained by total intragastric infusion. Journal of Animal Science 68, 2566–2574. Hammond, K.A. and Diamond, J.M. (1994) Limits to dietary nutrient intake and intestinal nutrient uptake in lactating mice. Phyiological Zoology 67, 282–303. Harmon, D.L. (1992) Dietary influences on carbohydrases and small intestinal starch hydrolysis capacity in ruminants. Journal of Nutrition 122, 203–210. Harmon, D.L., Avery, T.B., Huntington, G.B. and Reynolds, P.J. (1988) Influence of ionophore addition to roughage and high-concentrate diets on portal blood flow and net nutrient flux in cattle. Canadian Journal of Animal Science 122, 203–210. Hill, T.M., Schmidt, S.P., Russell, R.W., Thomas, E.E. and Wolfe, D.F. (1991) Comparison of urea treatment with established methods of sorghum grain preservation and processing on site and extent of starch digestion by cattle. Journal of Animal Science 69, 4570–4576. Holtug, K., Clausen, M.R., Hove, H., Christiansen, J. and Mortensen, P.B. (1992) The colon in carbohydrate malabsorption – short-chain fatty acids, pH and osmotic diarrhoea. Scandanavian Journal of Gastroenterology 27, 545–552. Huntington, G.B. (1982) Portal blood flow and net absorption of ammonia-nitrogen, ureanitrogen, and glucose in nonlactating Holstein cows. Journal of Dairy Science 65, 1155–1162. Huntington, G.B. (1983) Net nutrient absorption in beef steers fed silage or high concentrate diets containing four levels of limestone. Journal of Nutrition 113, 1157–1164. Huntington, G.B. (1997) Starch utilisation by ruminants: from basics to the bunk. Journal of Animal Science 75, 852–867. Huntington, G.B. and Prior, R.L. (1983) Digestion and absorption of nutrients by beef heifers fed a high concentrate diet. Journal of Nutrition 113, 2280–2288. Huntington, G.B. and Reynolds, P.J. (1983) Net volatile fatty acid absorption in non-lactating Holstein cows. Journal of Dairy Science 66, 86–92.

146

C.J. Seal and D.S. Parker Huntington, G.B., Reynolds, C.K. and Stroud, B.H. (1989) Techniques for measuring blood flow in the splanchnic tissues of cattle. Journal of Dairy Science 66, 1901–1908. Huntington, G.B., Eismann, J.H. and Whitt, J.M. (1990) Portal blood flow in beef steers: comparison of techniques and relation to hepatic portal blood flow, cardiac output and oxygen uptake. Journal of Animal Science 68, 1666–1673. Katz, M.L. and Bergman, E.N. (1969) A method for simultaneous catheterisation of the major splanchnic vessels of the sheep. American Journal of Veterinary Research 30, 655–661. Kelly, J.M., Vaage, A.S., McBride, B.W. and Milligan, L.P. (1989) Oxygen consumption and the energy costs of Na+, K+-ATPase in rumen epithelial papillae from Hereford steers. Journal of Dairy Science 72 (Suppl. 1), 560. Kim, E.J., Dhanoa, M.S., Gooden, J.M., Dawson, J.M., Buttery, P.J. and Scollan, N.D. (1998) The effect of dietary asynchrony on circadian changes in net nutrient fluxes across the portal-drained viscera in calves. In: McCracken, K., Unsworth, E.F. and Wylie, A.R.G. (eds) Proceedings of the 14th Symposium on Energy Metabolism, Newcastle, Co. Down, Northern Ireland. CAB International, Wallingford, UK, pp. 15–18. Krehbiel, C.R., Harmon, D.L. and Schneider, J.E. (1992) Effects of increasing ruminal butyrate on portal and hepatic nutrient flux in steers. Journal of Animal Science 70, 904–914. Kreikemeier, K.K., Harmon, D.L., Brandt, R.T.J., Avery, T.B. and Johnson, D.E. (1991) Effects of various levels of abomasal glucose, corn starch and corn dextrin on small intestinal disappearance and net glucose utilisation. Journal of Animal Science 69, 328–338. Kristensen, N.B., Danfaer, A., Tetens, V. and Agergaard, N. (1996) Portal recovery of intraruminally infused short-chain fatty acids in sheep. Acta Agriculturae Scandanavica. Section A. Animal Science 46, 26–38. Linington, M.J., Meyer, J.H.F. and Van der Walt, J.G. (1998a) Ruminal VFA production rates, whole body metabolic kinetics and blood hormone concentrations in sheep fed high- and low-fibre diets. South African Journal of Animal Science 28, 82–98. Linington, M.J., Meyer, J.H.F. and Van der Walt, J.G. (1998b) The use of energy by the splanchnic tissues and their metabolism of VFA, glucose and FFA in sheep fed high- and low-fibre diets. South African Journal of Animal Science 28, 99–119. Lobley, G.E. (1994) Nutritional and endocrine regulation of energy metabolism. In: Aguilera, J.F. (ed.) 13th Symposium on Energy Metabolism of Farm Animals, Mojacar, Spain. CSIC Publishing Service, Spain, pp. 139–149. Lobley, G.E., Milne, V., Lovie, J.M., Reeds, P.J. and Pennie, K. (1980) Whole body and tissue protein synthesis in cattle. British Journal of Nutrition 43, 491–502. Lobley, G.E., Connell, A., Revell, D.K., Bequette, B.J., Brown, D.S. and Calder, A.G. (1996a) Splanchnic-bed transfers of amino acids in sheep blood and plasma, as monitored through the use of a multiple U-13C-labelled amino acid mixture. British Journal of Nutrition 75, 217–235. Lobley, G.E., Weijs, P.J.M., Connell, A., Calder, A.G., Brown, D.S. and Milne, E. (1996b) The fate of absorbed and exogenous ammonia as influenced by forage or forage-concentrate diets in growing sheep. British Journal of Nutrition 76, 231–248. Lobley, G.E., Bremner, D.M., Nieto, R., Obitsu, T., Hotson Moore, A. and Brown, D.S. (1998) Transfers of N metabolites across the ovine liver in response to short-term infusions of an amino acid mixture into the mesenteric vein. British Journal of Nutrition 80, 371–379. López, S., Deb Hovell, F.D. and MacLeod, N.A. (1996) Osmotic pressure, water kinetics and volatile fatty acid absorption in the rumen of sheep, sustained by intragastric infusions. British Journal of Nutrition 71, 153–168. McBride, B.W. and Kelly, J.M. (1990) Energy costs of absorption and metabolism in the ruminant gastrointestinal tract and liver: a review. Journal of Animal Science 68, 2997–3010. MacRae, J.C., Bruce, L.A., Brown, B.S. and Calder, G. (1997) Amino acid use by the gastrointestinal tract of sheep given lucerne forage. American Journal of Physiology 273, G1200–G1207. Marconi, A.M., Battaglia, F.C., Meschia, G. and Sparks, J.W. (1989) A comparison of amino

Influence of Gastrointestinal Metabolism on Substrate Supply to the Liver

147

acid arteriovenous differences across the liver and placenta of the fetal lamb. American Journal of Physiology 257, E909–E915. Neutze, S.A., Oddy, V.H. and Gooden, J.M. (1994) A cranial mesenteric vein preparation for measurement of amino acid uptake by lambs. Journal of Agricultural Science, Cambridge 122, 309–314. Nocek, J.E. and Tamminga, S. (1991) Site of digestion of starch in the gastrointestinal tract of dairy cows and its effect on milk yield and composition. Journal of Dairy Science 74, 3598–3629. Okine, E.K., Cherry, R. and Kennelly, J.J. (1993) Glucose and amino acid transport and metabolism in duodenal flat sheets of dairy cattle at three stages of lactation. Comparative Biochemistry and Physiology 107, 719–726. Owens, F.N., Zinn, R.A. and Kim, Y.K. (1986) Limits to starch digestion in the ruminant small intestine. Journal of Animal Science 63, 1634–1648. Parker, D.S. (1990) Manipulation of the functional activity of the gut by dietary and other means (antibiotics/probiotics) in ruminants. Journal of Nutrition 120, 639–648. Pell, J.M., Caldarone, E.M. and Bergman, E.N. (1986) Leucine and ␣-ketoisocaproate metabolism and interconversions in fed and fasted sheep. Metabolism 35, 1005–1016. Perry, K.W., Janes, A.N., Weekes, T.E., Parker, D.S. and Armstrong, D.G. (1994) Glucose and L-lactate metabolism in pregnant and lactating ewes fed barley- or ground maize-based diets. Experimental Physiology 79, 35–46. Peters, J.P., Shen, R.Y.W., Robinson, J.A. and Chesters, S.T. (1990) Disappearance and passage of propionic acid from the rumen of the beef steer. Journal of Animal Science 68, 3337–3349. Piccioli Cappelli, F., Seal, C.J. and Parker, D.S. (1997) Glucose and [13C]leucine metabolism by the portal-drained viscera of sheep fed on dried grass with acute intravenous and intraduodenal infusions of glucose. British Journal of Nutrition 78, 931–946. Reeds, P.J. (1998) In: McCracken, K., Unsworth, E.F. and Wylie, A.R.G. (eds) Proceedings of the 14th Symposium on Energy Metabolism, Newcastle, Co. Down, Northern Ireland. CAB International, Wallingford, UK. Remond, D., Ortigues-Marty, I., Isserty, A. and Lefaivre, J. (1998) Technical note: measuring portal blood flow in sheep using an ultrasonic transit time flow probe. Journal of Animal Science 76, 2712–2716. Reynolds, C.K. and Huntington, G.B. (1988) Partition of portal-drained visceral net flux in beef steers. 1. Blood flow and net flux of oxygen, glucose and nitrogenous compounds across stomach and post-stomach tissues. British Journal of Nutrition 60, 539–551. Reynolds, C.K., Tyrell, H.F. and Reynolds, P.J. (1991) The effect of diet forage-to-concentrate ratio and intake on energy metabolism in growing beef heifers: net nutrient metabolism by gut tissues. Journal of Nutrition 121, 1004–1015. Reynolds, C.K., Harmon, D.L. and Cecava, M.J. (1994) Absorption and delivery of nutrients for milk protein synthesis by portal-drained viscera. Journal of Dairy Science 77, 2787–2808. Seal, C.J. and Parker, D.S. (1994) Effect of intraruminal propionic acid infusion on metabolism of mesenteric- and portal-drained viscera in growing steers fed a forage diet. 1. Volatile fatty acid, glucose and lactate. Journal of Animal Science 72, 1325–1334. Seal, C.J. and Parker, D.S. (1996) Effect of intraruminal propionic acid infusion on metabolism of mesenteric- and portal-drained viscera in growing steers fed a forage diet. 2. Urea, ammonia, amino acids and peptides. Journal of Animal Science 74, 245–255. Seal, C.J. and Reynolds, C.K. (1993) Nutritional implications of gastrointestinal and liver metabolism in ruminants. Nutrition Research Reviews 6, 185–208. Seal, C.J., Parker, D.S. and Avery, P.J. (1992) The effect of forage and forage-concentrate diets on rumen fermentation and metabolism of nutrients by the mesenteric and portal-drained viscera in growing steers. British Journal of Nutrition 67, 355–370.

148

C.J. Seal and D.S. Parker Seal, C.J., Parker, D.S., MacRae, J.C. and Lobley, G.E. (1994) Effect of intraduodenal starch infusion on net portal absorption of amino acids in growing steers fed lucerne. Animal Production 58, 434A. Sharp, W.M., Johnson, R.R. and Owens, F.N. (1982) Ruminal VFA production with steers fed whole or ground corn grain. Journal of Animal Science 55, 1505–1514. Shirazi-Beechey, S.P., Smith, M.W., Wang, Y. and James, P.S. (1991) Postnatal development of lamb intestinal digestive enzymes is not regulated by diet. Journal of Physiology 437, 691–698. Shoemaker, W.C. (1964) Methods and techniques for measurement of hepatic physiology and metabolism. In: Rouiller, C. (ed.) The Liver: Morphology, Biochemistry, Physiology. Academic Press, New York, pp. 243–266. Stangassinger, M. and Giesecke, D. (1986) Splanchnic metabolism of glucose and related energy substrates. In: Milligan, L.P., Grovum, W.L. and Dobson, A. (eds) Proceedings of the 6th International Symposium on Ruminant Physiology. Prentice Hall, Englewood Cliffs, New Jersey, pp. 347–366. Tagari, H. and Bergman, E.N. (1978) Intestinal disappearance and portal blood appearance of amino acids in sheep. Journal of Nutrition 108, 790–803. Taniguchi, K., Huntington, G.B. and Glenn, B.P. (1995) Net nutrient flux by visceral tissues of beef steers given abomasal and ruminal infusions of casein and starch. Journal of Animal Science 73, 236–249. Van der Walt, J.G., Baird, G.D. and Bergman, E.N. (1983) Tissue glucose and lactate metabolism and interconversions in pregnant and lactating sheep. British Journal of Nutrition 50, 267–280. Wangsness, P.J. and McGilliard, A.D. (1972) Measurement of portal blood flow in calves by dye dilution. Journal of Dairy Science 55, 1439–1446. Weekes, T.E.C. (1991) Hormonal control of glucose metabolism. In: Tsuda, T., Sasaki, Y. and Kawashima, R. (eds) Physiological Aspects of Digestion and Metabolism in Ruminants. Academic Press, New York, pp. 183–200. Weekes, T.E.C. and Webster, A.J.F. (1975) Metabolism of propionate in the tissues of the sheep gut. British Journal of Nutrition 33, 425–468. Whitt, J., Huntington, G., Zetina, E., Casse, E., Taniguchi, K. and Potts, W. (1996) Plasma flow and net nutrient flux across gut and liver of cattle fed twice daily. Journal of Animal Science 74, 2450–2461. Windmeuller, H.G. and Spaeth, A.E. (1978) Identification of ketone bodies and glutamine as the major respiratory fuels in vivo for postabsorptive rat small intestine. Journal of Biological Chemistry 253, 69–76. Yu, F., Bruce, L.A., Coop, R.L. and MacRae, J.C. (1998) Leucine metabolism across the gastrointestinal tract of sheep infected with Trichostrongylus colubriformis. Proceedings of the British Society of Animal Science p. 1. Zhao, F.-Q., Okine, E.K., Cheeseman, C.I., Shirazi-Beechey, S.P. and Kennelly, J.J. (1998) Glucose transporter gene expression in lactating bovine gastrointestinal tract. Journal of Animal Science, 76, 2921–2929.

9

The Liver: Integrator of Nitrogen Metabolism G.E. LOBLEY1, G.D. MILANO2 AND J.G. VAN DER WALT3 1Rowett

Research Institute, Bucksburn, Aberdeen, UK; 2Departamento de Fisiopatologia, Facultad de Ciencias Veterinarias UNCPBA, Campus Universitario, Paraje Arroyo Seco (7000) Tandil, Argentina; 3Department of Physiology, Faculty of Veterinary Sciences, University of Pretoria, Onderstepoort, South Africa

Introduction The liver is positioned at the anatomic and metabolic crossroads of the body and, as such, controls the amount and nature of nutrients available to the peripheral tissues from digestive tract absorption. This role in nutrient partitioning encompasses all of the major macronutrients and may be influenced by the involvement of the liver in the production (insulin-like growth factor-I, IGF-I) and removal (insulin, glucagon) of key hormones. For the sake of brevity, this review will focus on the metabolic products of protein-N, particularly ammonia and amino acids. Hepatic detoxification of the former, to prevent deleterious peripheral hyperammonaemia, is probably a more vital function in ruminants than non-ruminants, particularly for grazing animals. Because peripheral hyperaminoacidaemia may also produce adverse effects, it is evident that the regulation of systemic plasma amino acid concentrations is also important. This is achieved by modifying the quantity of individual absorbed amino acids removed by the liver and, in particular, controlling the amount oxidized versus the proportion converted to other metabolites, such as the plasma export proteins. Although the biochemical pathways involved in these partitions are well documented and a number of recent reviews have described the quantitative events (e.g. van der Walt, 1993; Reynolds, 1995; Lobley and Milano, 1997), the regulatory aspects still remain obscure. Furthermore, the limited data available to provide a conceptual framework for metabolic control mechanisms are often obtained from non-ruminants, or in vitro studies, and may not apply directly to ruminants. The shortcomings of these approaches should always be kept in mind.

Liver architecture The liver is the most richly vascularized organ in the body, with blood comprising approximately 25% of its mass. Despite constituting less than 2% of body weight in © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

149

150

G.E. Lobley et al.

adult ruminants, the liver also receives approximately 25% of cardiac output. Total blood flow through the liver is sensitive to nutrition, plus other regulatory factors, and can reach values of 3 l min⫺1 kg⫺1 liver weight (or 60 ml min⫺1 kg⫺1 body weight). Blood supply to the liver arises from two sources, the hepatic portal vein and the hepatic artery, although the latter makes only a small contribution to total hepatic blood flow in cattle and sheep (8–12%; Reynolds, 1995; Isserty et al., 1998). Regulation of hepatic blood flow is unique. The main blood supply, through the hepatic portal vein, is not controlled by the liver; even if vascular resistance is increased to a maximum (for example, by stimulation of the hepatic sympathetic nerves), the flow remains unaltered, although the pressure rises. Furthermore, the well-known autoregulation of arterial blood flow through most other organs (driven by the level of tissue oxygenation) is absent from the liver. Instead, the arterial flow appears to be inversely linked to the hepatic portal venous flow, providing the liver with a constant total flow (Lautt, 1996). The mechanism may involve steady release of adenosine into the Space of Mall, followed by removal through both arterial and hepatic portal inflows. If the portal supply is compromised, the local concentration of adenosine increases, stimulating arterial vasodilation and flow to remove the metabolite. During periods of maximal hepatic portal flow, e.g. at peak absorption from the rumen, the inverse should occur. While most information on spatial organization within the mammalian liver is derived from rodents, ruminants may differ somewhat. The liver is comprised of four cell types: hepatocytes (by far the largest group); macrophages (Kupffer cells); endothelial cells lining the Space of Disse; and fat storage (Ito) cells (see Gebhardt, 1992; Jungermann and Kietzmann, 1996). The hepatic acinus, accepted as the smallest functional unit, comprises parenchymal cells in berry-like clusters, approximately 2 mm in radius, on a vascular stalk (portal triad) containing the finest branches of the hepatic artery, hepatic portal vein and the bile duct. This arrangement means that blood flows past both sides of not more than 16 hepatocytes, before exiting via the central drainage system (Lautt, 1996). The hepatocytes exhibit heterogeneity in both spatial and biochemical characteristics and may be divided simplistically into two populations, periportal and perivenous (also known as pericentral), based on their ability to synthesize urea or glutamine (Fig. 9.1). Although other metabolic activities of these cells show a smooth transition along the acinus from periportal to perivenous (Jungermann and Kietzmann, 1996), the enzymes of the ornithine cycle and glutamine synthesis are sharply separated, at least in the rodent. Periportal cells are the first to receive blood from the afferent hepatic portal vein and contain characteristically all five enzymes of the ornithine (urea) cycle, plus glutaminase in the mitochondria. Furthermore, mRNA for a variety of export proteins, including albumin, are also present (Gebhardt, 1992; Jungermann and Kietzmann, 1996), as are transporters for most of the amino acids, with the notable exceptions of aspartate and glutamate. The perivenous cells are clustered around the efferent hepatic vein and lack the ornithine cycle enzymes, albumin mRNA and mitochondrial glutaminase. Instead, they possess cytosolic glutamine synthetase, the Xag transporter for glutamate and aspartate entry (Häussinger and Gerok, 1983), plus ornithine aminotransferase (necessary for the catabolism of arginine; Kuo et al., 1991). A similar perivenous localization of glutamine synthetase in sheep is suggested by the data obtained from antegrade and retrograde perfusion of isolated liver, with and without inhibition of glutamine synthesis by sulphoximine (Rossouw et al., 1997).

The Liver: Integrator of Nitrogen Metabolism

151

Blood flow

Amino acid catabolism

Periportal

Perivenous

Tyrosine aminotransferase

Serine dehydratase Ornithine aminotransferase

Ammonia and urea metabolism Carbamoyl phosphate synthetase Glutamine synthetase

Plasma protein synthesis Albumin, fibrinogen a-Fetoprotein, angiotensinogen

Fig. 9.1. Metabolic zonation and pattern of enzyme distribution between periportal and perivenous hepatocytes in non-ruminants (data from Jungermann and Kietzmann, 1996; O’Sullivan et al., 1998). The magnitude of the symbols indicates the relative contribution within each zone, e.g. serine dehydratase activity declines between periportal and perivenous hepatocytes, but still occurs in the latter, while glutamine synthetase is restricted to the perivenous region.

Ammonia metabolism and ureagenesis The spatial organization of hepatic cells confers a ‘metabolic zonation’ (Häussinger et al., 1992a), particularly with reference to ammonia metabolism and ureagenesis (Fig. 9.1). Based on the high Km of carbamoyl phosphate synthetase 1 (CPS1) for ammonia (approximately 1–2 mM; Lusty, 1978), the ornithine cycle provides a primary high capacity, low affinity system for the hepatic removal of ammonia. In contrast, limited amounts of perivenous glutamine synthetase, with a low Km for ammonia (0.11 mM in humans, Kaiser et al., 1988; 0.33 mM in rats, Deuel et al., 1978), offer a secondary, low capacity, high affinity system for capture of ammonia. The effectiveness of this two-stage system ensures that up to 98% of [15N]ammonia in the portal vein is removed during each pass across the liver (Nieto et al., 1996). This system is vital to maintain peripheral plasma ammonia concentrations at very low concentrations. Failure to do so may lead to a range of deleterious consequences, including reduced fertility

152

G.E. Lobley et al.

(McEvoy et al., 1997), lowered appetite and, in extreme cases, coma and death (Summerskill and Wolpert, 1970).

The ornithine cycle The ornithine cycle requires equal and concomitant inflows of N into carbamoyl phosphate and aspartate and, in theory at least, an imbalance at either entry point could lead to the potentially deleterious accumulation of metabolites (e.g. an inability to remove ammonia). Specific mechanisms exist to reduce the risk of such occurrences. For example, under normal cell redox conditions in periportal hepatocytes, the Keq of mitochondrial glutamate dehydrogenase (GDH) strongly favours synthesis of glutamate from ammonia and oxoglutarate. This action, when coupled with that of aspartate : 2-oxoglutarate transaminase (AOT), transfers ammonia-N to mitochondrial aspartate. This would be important when ammonia extraction by the liver exceeds hepatic catabolism of amino acids, probably the normal circumstance with ruminants fed fresh or conserved forages (e.g. Whitt et al., 1996; Lobley et al., 1998). Indeed, in studies involving [15N]ammonia across the sheep liver (Lobley et al., 1995; Milano et al., 1996) and with ovine hepatocytes (Luo et al., 1995; Mutsvangwa et al., 1997), [15N15N]urea was produced, proving that ammonia-N can be transferred to aspartate (Table 9.1). Furthermore, in ovine hepatocytes incubated with [15N]ammonia, the 15N enrichments of glutamate and aspartate were similar (Luo et al., 1995), confirming the involvement of GDH and AOT. The contribution of ammonia-15N to both carbamoyl phosphate and aspartate-N inflows can be obtained quite simply from isotopomer analysis of the relative proportions of 14N15N and 15N15N urea formed (Milano et al., 1996; Milano, 1997). For example, under conditions where rapid equilibrium occurs via GDH-AOT (Luo et al., 1995), the N precursor enrichment, a, is obtained from a = 2/[2 + (14N15N : 15N15N)] Thus, under these circumstances, only a sample of urea is required to determine this contribution. A similar approach has been applied to studies with the perfused rat liver (Brosnan et al., 1996). More sophisticated equations have been developed (Milano, 1997) to accommodate situations where equilibrium between ammonia and aspartateN is not achieved, as is the probable case in vivo (Lobley et al., 1995). This approach has shown that in fasted sheep, subjected to an ammonia overload, approximately 32% of aspartate-N was derived from ammonia (Milano et al., 1996). When coupled with net transfers of ureagenic metabolites across the liver, this technique offers a powerful means to determine how much of the urea truly arises from ammonia- or amino acid-N. In non-ruminants, or ruminants offered grain-based rations with additional rumen undegradable protein, supply of amino acid-N may exceed that of ammonia (Burrin et al., 1991; Goetsch and Ferrell, 1995), and extra ammonia must be generated to maintain ornithine cycle activity and prevent peripheral hyperaminoacidaemia. This is achieved through mitochondrial glutaminase, which generates ammonia (and glutamate) from glutamine. Indeed, glutaminase may act as a continuous amplification mechanism to maintain high concentrations of ammonia within the mitochondria (Meijer et al., 1985), to compensate for the high Km for ammonia of CPS1. In sheep,

The Liver: Integrator of Nitrogen Metabolism

153

with intra-abomasal infusions of either ammonium bicarbonate or amino acids, coupled with intravascular infusion of 5-15N (amido) glutamine, between 28 and 61% of the [15N]urea produced across the liver was derived directly from glutamine (R.M. Nieto, T. Obitsu, A. Fernandez and G.E. Lobley, unpublished results). Net hepatic removal of glutamine in pregnant, dry cows was proportional to the amounts of an amino acid mixture infused into the mesenteric vein (Wray-Cahen et al., 1997). In contrast, complementary studies involving amino acid infusions into sheep (Lobley et al., 1998) failed to alter liver glutamine uptake. These observed differences probably relate to the proportions of ammonia and amino acids extracted by the liver (Lobley et al., 1998). Amino acid-N may also transfer to ammonia more directly, because ovine hepatocytes, incubated with unlabelled ammonia and [15N]alanine, also produced [15N15N]urea, although this was smaller than the amount of [14N15N]urea synthesized (Mutsvangwa et al., 1997; Table 9.1). Whether these various mechanisms operate effectively enough may have important implications for ruminant productivity. Before the paradigm of the ornithine/glutamine cycles was proposed, it was thought that, under conditions of ammonia excess, either peripheral hyperammonaemia would occur or amino acid catabolism would be stimulated to provide additional aspartate-N. Indeed, from retrospective analysis of transhepatic ruminant data, a scenario was proposed that removal of ammonia-N by the liver was accompanied by a similar input from other N sources, i.e. amino acid catabolism would match ammonia detoxification (Parker et al., 1995). This hypothesis would

Table 9.1. Effect of propionate on urea production in ovine hepatocytes and sources of ornithine cycle-N. nmol mg⫺1 wet cells h⫺1† Propionate* (mM) 0 0.2 0.4 0.8 1.6

Propionate‡ (mM) 0 0.31 0.63 1.25

NH3 uptake 17a 58b 54b 43c 36c

Urea production

Glucose production

Urea ([15N15N]:[14N15N])

28 63b 62b 54c 43c

6a 16b 22b 29c 31c

1.0 1.1 1.1 1.2 1.1

nmol mg⫺1 DM h⫺1† Urea production 7.1a 6.9a 7.2a 3.9b

Urea ([15N15N]:[14N15N]) 0.29 0.31 0.26 0.27

* Medium also contained 0.67 mM 15NH4Cl, plus a physiological mixture of amino acids; data from M.A. Lomax and G.E. Lobley (unpublished data). † Within study different superscripts in a column indicate P < 0.05 or better. ‡ Medium also contained 0.63 mM NH Cl, plus 1.25 mM [15N]alanine; data from Mutsvangwa et al. (1997). 4

154

G.E. Lobley et al.

apparently account for the poorer utilization of dietary amino acids by ruminants, particularly those fed fresh or conserved forages. In the event, controlled perturbation of ammonia supply to the liver, by intra-mesenteric vein infusion of ammonium bicarbonate, showed that the extra urea produced exceeded ammonia uptake by between 13 and 17% (Lobley et al., 1996b; Lobley and Milano 1997). This is substantially less than the doubling predicted by the hypothesis but, none the less, does represent increased amino acid catabolism – thereby losing the equivalent of 0.6–1.3 g amino acid-N for daily anabolism. With ovine hepatic preparations in vitro, addition of ammonia led to similar amounts of N released as urea plus glutamine (Luo et al., 1995; Rossouw et al., 1997, 1999), i.e. the removal of ammonia did not appear to stimulate amino acid catabolism. Overall, therefore, the importance of ammonia detoxification may well have resulted in the process being effectively uncoupled from amino acid oxidation.

Capacity and substrate priority for the ornithine cycle There is a finite capacity for the rate of ureagenesis and a number of studies in both cattle and sheep have yielded values of 2 mmol urea-N kg⫺1 fresh liver weight min⫺1 (see Lobley et al., 1998). This amounts to approximately 29 and 435 g urea-N day⫺1, for a 40 kg sheep or a 600 kg bovine respectively, within the range of N intake by these animals on high levels of nutrition (Sarraseca et al., 1998). At higher intakes, unless ornithine cycle enzyme content or activity increases adaptively, amino acids and ammonia will escape beyond the liver. Furthermore, even at lower mean daily urea production rates, diurnal fluctuations in ammonia and amino acid absorption (Whitt et al., 1996) may result in periods when these capacities are exceeded. This raises the important question: when the capacity of the ornithine cycle is reached does the liver prioritize removal of ammonia or of amino acids? When ammonia inflow to the liver of young sheep was raised from 0.58 to 2.01 mM min⫺1, through acute (31 min) infusions of ammonium bicarbonate into the mesenteric vein, hepatic extraction was elevated from 0.44 to 1.21 mM min⫺1 (Milano, 1997; see Lobley and Milano, 1997). This incomplete removal led to ammonia overspill into the peripheral circulation and elevated arterial concentrations (from 0.08 to 0.31 mM), a finding compatible with the low capacity of glutamine synthetase in the ovine perivenous hepatocytes (Rossouw et al., 1999). Interestingly, the nonammonia N contribution to ureagenesis declined (from 0.36 to 0.14 mM min⫺1; P = 0.07), suggesting that amino acid catabolism was reduced in order to allow more ammonia to be detoxified. These acute ammonia infusions might relate to situations of diurnal feeding, e.g. morning or evening intakes of fresh grass at pasture, where it is possible that absorption of large quantities of ammonia might actually reduce, temporarily, the hepatic oxidation of amino acids. Interestingly, this is the converse of the original ammonia hypothesis (Parker et al., 1995). This priority in favour of ammonia removal was confirmed in a converse study when 1.84 mM min⫺1 of mixed amino acid-N was infused into the mesenteric vein of sheep fed a basal supra-maintenance grass pellet ration (Lobley et al., 1998). Even though the potential ureagenic supply to the liver exceeded ornithine cycle activity, hepatic ammonia removal was unchanged and severe arterial blood hyperaminoacidaemia resulted. Although hepatic extraction favours ammonia removal, neither substrate was able to completely suppress in vivo the

The Liver: Integrator of Nitrogen Metabolism

155

ureagenic capacity of the other. This is despite the observations with ovine hepatocytes in vitro that ureagenesis can proceed with ( 15N) ammonia as the sole exogenous N substrate and with [ 15N15N]urea as the primary product (Luo et al., 1995); a finding confirmed with the perfused sheep liver (Rossouw et al., 1997, 1999).

Control of ureagenesis A major aim of any high-production system is to increase the amount of anabolic metabolites (e.g. amino acids) available to the peripheral tissues. Clearly, any treatment that changes anabolism without changing intake or absorption of nutrients must alter ureagenesis (e.g. bovine-somatotropin; Bruckentahl et al., 1997). Despite many years of research, it is still unclear whether ureagenesis is actively or passively controlled. Simple consideration of the quantities and maximal activities of the ornithine cycle enzymes would suggest that arginosuccinate synthetase (ASS) is limiting (Rattenbury et al., 1980). In practice, however, CPS1 is a more probable candidate because this is positively regulated by the mitochondrial concentration of N-acetylglutamate (N-AG; see Meijer et al., 1990). The activity of N-AG synthetase is regulated in vitro by several factors, including insulin (an inhibitor) and glucagon (an activator). This would fit with the competing roles of these hormones; glucagon leads to poorer N retention (thus more urea synthesis), while insulin diverts amino acids towards peripheral tissue anabolism. Direct, or immediate, actions of insulin on the ornithine cycle, however, are probably doubtful in vivo. For example, although the liver of cattle, which were acutely (3 h) hyper-insulinaemic, extracted substantial amounts of insulin, urea production remained unchanged (Eisemann and Huntington, 1994). In other bovine studies, chronic administration of growth hormone (or its releasing factor) decreased hepatic amino acid extraction and ureagenesis (Bruckentahl et al., 1997), although this has not been a universal finding (McLeod et al., 1997). The long-term nature of these latter studies does not allow resolution of direct hormonal effects on the ornithine cycle enzymes. Another putative regulator of ureagenesis is propionate, which has been shown at supra-physiological concentrations to be a potent in vitro inhibitor of ureagenesis (e.g. Stewart and Walser, 1980). Such regulation through propionate would provide an attractive explanation for the superior performance of animals on concentrate versus forage diets. This role has been reinvestigated recently with ovine hepatocytes and, while a minimal amount of propionate (or other 3-carbon unit) is needed to activate the ornithine cycle, ureagenesis is suppressed at propionate concentrations above 0.63 mM (Mutsvangwa et al., 1997; see Table 9.1). At these higher propionate concentrations, gluconeogenesis continues to be stimulated, thus breaking the normal link between glucose synthesis and urea production. Inhibition of ureagenesis by propionate probably does not affect the relative inflows into the ornithine cycle of ammonia and amino acids, as shown by incubating hepatocytes in the presence of either [15N]ammonia or [15N]alanine. Indeed, the relative ratios of [15N15N]:[14N15N] urea produced were uninfluenced by the presence of propionate (Table 9.1). While these effects are only apparent towards the upper end of the physiological range in vitro, the elevated arterial ammonia concentrations, which result from episodic infusion of propionate into the rumen (Choung and Chamberlain, 1995), may indicate that the ornithine cycle may also be inhibited in vivo.

156

G.E. Lobley et al.

Fate of urea synthesized In ruminants, urea-N synthesis often exceeds the amount of apparently digestible N and, at submaintenance intakes, even total ingested N (e.g. Goetsch and Ferrell, 1995; Sarraseca et al., 1998). This apparent anomaly is due to substantial entry of urea into the gastrointestinal tract (GIT), where it may either provide an N source for microbial protein synthesis (and thus contribute to potential N anabolism) or be re-absorbed as ammonia across the portal-drained viscera (PDV) (with augmentation of hepatic ureagenesis). Hitherto, resolution of these two fates has been difficult but can now be overcome by the use of the [15N15N]urea isotopomer technique (see p. 152, Fig. 9.2). This method estimates in normal humans that approximately 80–90% of the urea-N which enters the GIT may be returned for anabolic purposes. This proportion may be markedly influenced by dietary treatments (Jackson, 1998). Hitherto, this approach has failed to account for multiple re-entries of [14N15N]urea into the GIT, leading to the formation of daughter [14N15N]molecules, indistinguishable chemically from the parent labelled urea. This leads to over-estimates of the [14N15N]urea transfers and, for ruminants, the error introduced can be as large as 60%. Application of this isotopomer technique to sheep, but with correction for multiple entries of [14N15N]urea to the GIT, revealed that although absolute rates of

Fig. 9.2. Re-utilization of hepatic urea from the digestive tract quantified by use of infusion of [15N15N]urea. Urea-N is either lost in faecal-N, converted to microbial protein (then digested and absorbed as amino acids), or reabsorbed as 15NH3 when it can either be reconverted to urea (as 14N15N) or used to aminate (transaminate) amino acids.

The Liver: Integrator of Nitrogen Metabolism

157

ureagenesis and urea-N entry to the GIT doubled as intake of grass pellets was increased from 0.6 to 1.8 ⫻ maintenance, the proportion recycled to the ornithine cycle remained constant and represented 22–30% of urea production (Sarraseca et al., 1998). Similar values (30–31%) were obtained with rations of either a mixed forageconcentrate or chopped hay plus grass pellets (unpublished results). In practice, this means that, under these dietary conditions (Lobley et al., 1996b, 1998), as much as 50% of the ammonia absorbed from the GIT and detoxified by the liver arises from urea synthesized earlier and thus represents part of N recycling in the ruminant.

Amino acid metabolism Liver N balance Studies that have examined trans-hepatic nitrogen metabolism are confounded by a number of apparent anomalies. Logically, there must be a sensible balance between N substrate inflow and outflow, with due allowance made for synthesis of export (plasma) proteins and accretion (or depletion) of liver protein mass. In practice, several recent studies show larger outflow of urea-N than the combined inflow of ammonia- and amino acid-N, even at high intakes, and before allowance is even made for net protein synthesis (Burrin et al., 1991; Reynolds and Tyrrell, 1991; Goetsch and Ferrell, 1995). Some studies found both negative and positive balances (e.g. Reynolds et al., 1994; Bruckentahl et al., 1997; Wray-Cahen et al., 1997) while others reported only net positive N uptake (e.g. Lobley et al., 1995, 1996b, 1998; Krehbiel et al., 1998). To blame may be unsatisfactory methodology, i.e. ‘␣-amino N’ (ninhydrin based), ‘free amino N’ (trinitrobenzene sulphonate) techniques, since these data may be up to threefold higher than values obtained from a full amino acid analysis (Burrin et al., 1991; compare Reynolds and Tyrrell, 1991 with Reynolds et al., 1994). Of course, some of the discrepancies may be accounted for by other N inputs, notably peptides (Webb, 1999). In many situations, however, physiologically-sensible balances can be achieved using improved techniques to measure the main N substrates (ammonia, urea and free amino acids), without involving peptide uptake. As a result, peptide exchanges are probably minor (but not necessarily unimportant). Further precision can be gained by a gravimetric approach (Lobley et al., 1998) or by a stable-isotope dilution technique (Calder et al., 1999).

Plasma versus blood transport Amino acids may be transported in either the plasma and/or the red blood cells. In sheep (Heitmann and Bergman, 1980) and cattle (Houlier et al., 1991), earlier studies on mass exchanges suggested that carriage to the liver was via plasma, but that hepatic transfer to peripheral tissues involved the erythrocytes. The former concept was supported, in sheep, by data with [13C] amino acids infused systemically and where, across the PDV, absorption caused a dilution of enrichments in the plasma but not of those in the red blood cell (Lobley et al., 1996a). Across the liver, however, there was only a small dilution of plasma enrichments (0–6%), suggesting that transfers were primarily

158

G.E. Lobley et al.

unidirectional, i.e. into the liver cells. Furthermore, because liver intracellular enrichments usually exceed those in the red blood cell (Lobley et al., 1996a; Connell et al., 1997), then any outward transfer from hepatocytes to erythrocytes should increase enrichment in the latter. In practice, enrichment within the red blood cells was either unchanged or decreased slightly (0 to ⫺6%). These data suggest strongly there is little net transfer of amino acids from the liver via the erythrocytes. This issue needs to be resolved, because of the marked consequences for both net and gross (isotopic) transfers.

Fates of individual amino acids For each individual amino acid, the proportion of that absorbed from the PDV, which is then removed across the liver, varies widely (see Table 9.2), although there are some consistencies. For example, the branch-chain amino acids (BCAA) have the lowest rates of fractional removal, under a wide range of physiological and nutritional conditions, in both cattle and sheep. Post-hepatic supply of BCAA will thus closely reflect changes in intake and absorption. The capacity of the liver to degrade BCAA is limited and, once this is exceeded, additional inflows are not extracted and catabolism proceeds in the extra-hepatic tissues (Lobley, 1992). In a number of studies with cattle and sheep, the net splanchnic appearance of phenylalanine has either been negative (Burrin et al., 1991; Koeln et al., 1993), close to zero (Reynolds et al., 1994), or only slightly positive (Lobley et al., 1995, 1996b; Bruckentahl et al., 1997; Table 9.2). In sheep, a similar situation exists for histidine (see Table 9.2). Some of these variations may again relate to analytical difficulties, because the proportion of total hepatic inflow that is removed may be as low as 1% (Table 9.3). Indeed, it has been argued that expression of the hepatic extraction against total amino acid inflow (hepatic portal plus hepatic artery flows) has greater physiological validity than comparison with net absorption (e.g. Reynolds et al., 1994; Hanigan et al., 1998). This approach does ‘dampen’ the differences between studies (compare lysine in Tables 9.2 and 9.3, for example) and may have particular use in the development of modelling equations (Hanigan et al., 1997). Notwithstanding analytical uncertainties, hepatic fractional extractions of specific amino acids (Tables 9.2 and 9.3) do vary in response to differences in total intake (Burrin et al., 1991), protein and b-ST supply (Bruckentahl et al., 1997) and amino acid infusions (Reynolds et al., 1994; Wray-Cahen et al., 1997; G.E. Lobley, unpublished data). This variation may relate, in part, to the physiological role of the liver as the site of synthesis of vital metabolites, e.g. glucose, neurotransmitters, xenobiotic conjugates, transmethylation products, etc., all of which involve the use of carbon skeletons derived from amino acids. For example, removal of glycine by the liver is associated with one-carbon metabolism and the synthesis of hippuric acid. There are finite demands for both of these and, when satisfied, glycine uptake by the liver is reduced. In addition, hepatic export proteins are relatively rich in phenylalanine, histidine and cysteine (64, 41 and 66 g kg⫺1 bovine albumin respectively versus 36, 23 and 26 g kg⫺1 for non-wool whole body sheep protein; MacRae et al., 1993). The pattern of amino acids extracted by the liver to fulfil these various functions does not match that absorbed from the PDV. Consequently, post-splanchnic supply of amino acids to the peripheral tissues differs considerably from that either absorbed from the PDV or

Cattle

Lapierre Koeln et al. et al. (2000) (1993)a Alanine Arginine Glycine Serine Tyrosine Histidine Isoleucine Leucine Lysine Methionine Phenylalanine Threonine Valine

0.41 0.25 0.67 0.52 0.50 0.39 0.01 0.20 0.14 0.29 0.64 0.23 ⫺0.04

0.58 2.66 1.18 0.80 1.16 ⫺0.13 ⫺0.01 0.23 0.53 1.12 1.11 0.56 0.32

Reynolds et al. (1994)b Control 1.17 – 1.00 1.10 0.90 0.57 0.45 0.45 0.71 – 1.00 0.84 0.50

Sheep Wray-Cahen et al. (1997)c

Bruckentahl et al. (1997)d

Burrin et al. (1991)e M

+Ala

Pre

+AA

Control

+Casein

0.75 – 0.82 0.85 0.80 0.50 ⫺0.08 0.13 0.41 – 0.82 0.48 0.17

0.97 0.20 1.26 0.73 – 0.57 0.41 0.01 0.16 0.70 0.67 0.72 0.12

0.54 0.70 0.73 0.37 – 0.50 0.49 0.30 0.31 0.83 0.87 0.49 0.25

0.70 0.42 0.70 1.12 0.57 0.61 0.40 0.35 0.42 0.68 0.94 1.79 0.45

0.74 0.80 0.60 0.46 0.78 0.78 0.32 0.28 0.45 0.73 0.84 1.32 0.41

1.30 ⫺25.5 1.29 0.49 0.89 ⫺0.85 ⫺0.23 ⫺0.14 ⫺1.94 ⫺0.36 0.49 0.43 ⫺0.27

Lobleyf

AL

Control

+AA

3.18 7.29 2.01 1.65 1.46 2.01 0.50 0.42 0.44 1.37 1.39 1.04 0.63

0.92 0.38 2.43 0.61 0.89 1.64 0.28 0.43 0.68 0.83 1.03 0.24 0.25

0.66 0.39 0.88 0.44 – 0.83 0.02 0.08 0.38 0.61 0.63 0.37 ⫺0.08

The Liver: Integrator of Nitrogen Metabolism

Table 9.2. Fractional extraction by the liver of absorbed amino acids.

a Fed

calves. 75% maize:soybean meal, with (+ala) or without (control) infusion of L-alanine into the mesenteric vein. c Pregnant dry cows (plasma data), basal ration (pre) plus intra-mesenteric vein infusion of an amino acid mixture (+AA). d Fed a high concentrate diet without (control) or with (+casein) 300 g day⫺1 casein infused into the abomasum. e Sheep at maintenance (M), or fed ad libitum (AL). f Sheep fed at 1.2 ⫻ maintenance with (+AA) or without (control) intra-mesenteric vein infusion of amino acid mixture (equivalent to 4 ⫻ basal intake); unpublished data. b Fed

159

160

G.E. Lobley et al.

Table 9.3. Per cent hepatic extraction of total amino acid inflows to the livera. Cattle

Alanine Arginine Glycine Serine Tyrosine Histidine Isoleucine Leucine Lysine Methionine Phenylalanine Threonine Valine

Sheep

Controlb

+alab

Controlc

Controld

+AAd

19.7 – 8.8 16.7 11.2 3.9 3.1 3.3 6.3 – 13.3 9.1 0.7

19.6 – 6.4 15.1 12.4 2.5 0.8 1.5 5.2 – 13.6 7.4 1.0

6.7 1.8 7.2 9.4 8.2 3.0 1.7 0.1 1.7 8.8 5.9 7.0 0.5

9.7 6.4 3.4 9.3 7.5 5.7 2.3 2.4 4.3 9.9 11.5 1.2 1.2

13.0 6.0 8.7 18.3 – 11.0 1.4 1.4 5.4 14.4 12.3 7.3 1.0

a Based

on portal vein + hepatic artery inflows and hepatic vein outflows. et al. (1994); fed 75 maize-soybean meal with (+ala) and without (control) alanine infusion (3 days). c Wray-Cahen et al. (1997); pre-infusion data only. d G.E. Lobley; unpublished results, with (+AA) and without (control) mesenteric vein infusion (4.5 h) of amino acid (AA) mixture (equivalent to 4 ⫻ basal intake). b Reynolds

required by peripheral tissues. In sheep, for example, this may result in a relative underprovision of histidine and phenylalanine, with a lesser deficit of methionine. Their availability may limit peripheral tissue growth, but leave the other essential amino acids in excess and subject to catabolism. If, however, plasma export proteins can provide amino acids for peripheral tissue needs, then, based on known rates of synthesis (Connell et al., 1997), the limitation changes to methionine and, to a lesser extent, phenylalanine. This fits with the generally perceived view that methionine is often the first-limiting amino acid for many ruminant diets. Extra-hepatic use of hepatic export proteins has been shown in rodents, by adding serum proteins tagged with residualizing labels (Thorpe et al., 1993), where as much as 20% of albumin is degraded in each of skeletal muscle and skin (Maxwell et al., 1990). Interestingly, hepatic removal and degradation of albumin appears to be dominated by endothelial cells (Eskild et al., 1989), adding another aspect to liver metabolic zonation. Because the balance between the synthesis of metabolites and export protein synthesis will vary with physiological state, then clearly neither the magnitude of hepatic fractional extractions nor the pattern of amino acids available to the periphery will be constant. In sheep, protein intake and parasitic infestation (Abbott et al., 1985) markedly influence both albumin pool size and biological half-life (degradation and clearance). Ovine albumin synthesis is sustained even under fasting conditions (Koeln et al., 1993) and is sensitive to intake (Connell et al., 1997), at least up to a certain limit of amino acid supply (Lobley et al., 1998).

The Liver: Integrator of Nitrogen Metabolism

161

Glutamine Aside from the major metabolic role of glutamine, it is now used as a therapeutic aid in certain clinical situations, particularly those where cells proliferate, such as in the GIT epithelia and the immune system (Heitmann and Bergman 1978; Calder 1995). Besides a key role in maintaining ureagenic flux and ammonia detoxification, glutamine may also impact on hepatic mechanisms related to acid–base balance and regulation of protein turnover (see next section). Maintenance of acid–base balance is a major role for the liver in non-ruminants. In rodents, chronic metabolic acidosis, induced by either HCl or ammonium chloride, results in suppression of ureagenesis and urinary urea elimination, but with concomitant stimulation of amino acid oxidation, net glutamine synthesis and urinary NH4+ release (Welbourne et al., 1986; May et al., 1992). Production of urea yields a neutral product (involving utilization of both NH4+ and HCO3⫺), whereas synthesis of glutamine spares bicarbonate. This bicarbonate is then available to neutralize protons, and, on transfer to the kidney, the amido group is released into urine as ammonium (NH4+) ion (Meijer et al., 1990). In sheep, acidosis induced by ammonium chloride had little effect on whole body and liver flux of [U-14C]glutamine (Heitmann and Bergman, 1978), but the complication of acidosis and the additional supply of free ammonia confounded interpretation. Certainly, mild acidosis, induced by [15N]ammonium chloride infusion into sheep, resulted in elevated enrichments of hepatic [15N]glutamine, comparable to those observed for urea (Lobley et al., 1995). Net movements of glutamine across the liver are little affected by ammonia but, under both neutral and acidotic conditions, there is a large change in rates of both utilization and synthesis (Table 9.4), i.e. transfer of ammonia through the glutamine amido group is stimulated. For sheep infused with HCl, urea production and urinary elimination were elevated, with a smaller increase in hepatic glutamine synthesis (Milano, 1997). The latter Table 9.4. Effect of ammonia supply and acidosis on ovine hepatic glutamine kinetics. Glutamine kinetics (␮mol min⫺1) Net utilization

Hepatic synthesis

Hepatic utilization

Study 1 Control +NH4Cl

⫺30a ⫺11b

47a 87b

⫺77a ⫺98b

Study 2 Control +HCl

⫺43 ⫺37

26 42

⫺70 ⫺79

Study 3 Control +NH4HCO3

⫺21 ⫺33

37a 51b

⫺58 ⫺84

Data from: Study 1, [U-14C]glutamine, acidosis, Heitmann and Bergman (1980); Study 2, [5-15N]glutamine, acidosis, Milano (1997); Study 3, [5-15N]glutamine, ammonia overload (R.M. Nieto, T. Obitsu, A. Fernandez, G.E. Lobley, unpublished results). For each study, values within a column but with different superscripts are significantly different by P < 0.05 or lower.

162

G.E. Lobley et al.

accounted for a minor fraction of the extra NH4+ appearing in the urine, with most arising from extra-hepatic sources. This suggests that the liver is probably less important in regulation of acid–base balance in sheep, compared with the rat (Fig. 9.3). For example, the activity of the ovine hepatic glutamine synthetase (50–100 ␮mol min⫺1 kg⫺1 liver) is only 15% that of the rat (Häussinger, 1990). Thus, the maximal rate of urea production in sheep exceeds hepatic glutamine synthesis by 5–10-fold (Rossouw et al., 1999; Table 9.5), compared with equal activities in the rodent. Furthermore, both the ornithine cycle and glutamine synthesis are sensitive to pH in the rat but not in the sheep, as shown by perfusion of the ovine dorsal lobe with buffers varying in pH from 7.2 to 7.6 (Rossouw et al., 1999; Table 9.5). There are other metabolic differences in acid–base responses between species. Notably, net changes in protein-N retention appear to be regulated by decreased protein synthesis in sheep (Milano, 1997) but by elevated protein degradation in rodents (Manier et al., 1994; Fig. 9.5).

Hepatic protein turnover Liver protein metabolism has two major components – production, or replacement, of constitutive material and synthesis of proteins destined for export. The liver responds rapidly to alterations in intake with short-term, but limited, fluctuations in protein mass (Lobley, 1990). These short-term changes may form part of a small, but mobile, protein reserve, sensitive to acute variations in protein supply (MacRae et al., 1991). Liver protein mass also alters, both in absolute terms and relative to body weight, in response to longer-term changes in nutrition (Burrin et al., 1992; Lobley et al., 1994). The rate of change is relatively slow, however, even when intake is altered from

Protein synthesis

Protein synthesis

=

Protein degradation ≠

WHOLE BODY

Ø

Protein degredation =

Protein oxidation



Protein oxidation



Urea synthesis

Ø

Urea synthesis



Net glutamine



Net glutamine

=(≠)

Urea



Urea



NH4+



NH4+



Fig. 9.3. Responses in whole body protein metabolism, plus liver and kidney urea and glutamine kinetics to acidosis in rodents and sheep (from Welbourne et al., 1986; May et al., 1992; Manier et al., 1994; Milano, 1997).

The Liver: Integrator of Nitrogen Metabolism

163

Table 9.5. Effect of ammonia supply and acidosis on ovine hepatic glutamine kinetics in vitro. Condition

NH3 uptake

Urea production

Glutamine production

Antegrade +NH4Cl (0.3 mM), pH 7.2 +NH4Cl (0.3 mM), pH 7.4 +NH4Cl (0.3 mM), pH 7.6 ⫺NH4Cl, pH 7.4

827 ± 134 770 ± 118 791 ± 96 ⫺24 ± 40a

596 ± 66 590 ± 71 594 ± 60 34 ± 10a

176 ± 27 172 ± 40 184 ± 34 40 ± 10a

Retrograde +NH4Cl (0.3 mM), pH 7.2 +NH4Cl (0.3 mM), pH 7.4 +NH4Cl (0.3 mM), pH 7.6 ⫺NH4Cl, pH 7.2

606 ± 83 707 ± 108 730 ± 83 ⫺19 ± 20a

530 ± 52b 590 ± 58 556 ± 60 38 ± 40a

192 ± 19 222 ± 26c 186 ± 28 20 ± 4a

Values are expressed as nmol N g⫺1 wet weight liver min⫺1 and are the means of seven perfusions (H. Rossouw, J.G. van der Walt, unpublished results). All analyses conducted within direction of flow. a Differs from all values obtained with NH Cl, at all pH values (P < 0.05). 4 b Differs from the value obtained with NH Cl, at pH 7.4 (P < 0.10). 4 c Differs from the value obtained with NH Cl, at pH 7.6 (P < 0.10). 4

30 g kg⫺1 weight day⫺1 to either ad libitum or just sufficient to maintain body weight. Under these conditions, fractional rates of protein gain or loss were, at maximum, 5 and ⫺1.5% day⫺1 (Burrin et al., 1992). In contrast, fractional rates of hepatic protein synthesis are much greater, but are also very variable (from 3 to 98% day⫺1; see Lescoat et al., 1997). This wide disparity makes interpretation of metabolic regulation difficult, with the problem exacerbated by technical difficulties. These include the under-estimation of synthesis of export proteins, as well as those constitutive proteins with high rates of turnover, using the continuous infusions of labelled amino acids. This problem is further complicated by the presence of different precursor pools within the liver. Mixed intracellular sources provide for the synthesis of constitutive proteins, while polypeptides destined for secretion are derived from amino acids that have newly entered the cell (see Connell et al., 1997). Notwithstanding these technical concerns, the large difference between fractional rates of protein synthesis and gain indicates that constitutive proteins are extensively degraded within the ruminant liver. If these liberated amino acids were to be channeled into a pool for catabolism, it would significantly affect the overall N economy of the animal. There is also strong evidence that the regulation of constitutive protein production differs from that of proteins destined for export. For example, in the transition from fasting to feeding, synthesis of albumin is markedly stimulated, in both absolute and fractional terms (Connell et al., 1997). Regulation through synthesis probably also largely accounts for changes in plasma protein concentration and content commonly observed under conditions of stress (e.g. Abbott et al., 1985) and under-nutrition (e.g. Liu et al., 1995). In contrast, liver constitutive protein synthesis is less susceptible to acute changes in amino acid supply (Lobley et al., 1998) or general intake (Connell et al., 1997). Consequently, changes in ovine hepatic protein mass probably occur through alteration in protein catabolism, possibly involving up- and down-regulation of lysosomal activity as proposed for the perfused rodent liver (Mortimore et al., 1989).

164

G.E. Lobley et al.

How then is such control exerted? Recent studies with rodent hepatocytes and perfused livers have indicated that the mechanism probably involves changes in cell volume (Häussinger et al., 1994). Cellular uptake of a number of amino acids, mediated by Na+-dependent transporters (e.g. the A system for glycine and alanine, or the Nsystem in the case of glutamine), is accompanied by concomitant inflow of Na+. This is then expelled by exchange with K+ ion, the energy being provided by the membrane Na+-K+ ATPase. This flow of ions leads to increased uptake of intracellular water and, therefore, cell swelling. As cell volume expands, protein degradation is reduced, probably related to a decrease in the rate of formation and acidification of lysosomes (Luiken et al., 1996). Under the converse conditions, i.e. efflux of these amino acids, the hepatocytes shrink, degradation is enhanced and protein synthesis inhibited (Häussinger et al., 1994). Net production of export protein appears to be regulated primarily by changes in synthesis, whereas constitutive protein mass is modulated more by alterations in the rate of proteolysis (Connell et al., 1997). These contrasting mechanisms may have a common step involving intracellular targetting of the mRNAs for the export proteins to the endoplasmic reticulum (Hesketh et al., 1998) and phosphorylation of ribosomal protein S6 (Blommart et al., 1995). The binding there of additional ribosomes will augment synthesis of export proteins and, at the same time, reduce the release of reticulum membrane needed for formation of the autophagosomes. Thus, proteolysis would be inhibited. If targetting and binding to the endoplasmic reticulum is inhibited (as at low intakes) then export protein synthesis decreases while lysosome formation and activity is enhanced. Interestingly, two major protein anabolic hormones, insulin and IGF-I, also lead to increased cell hydration, while the catabolic hormone, glucagon, causes shrinkage. Both insulin and glucagon are extracted by the ruminant liver (Lapierre et al., 1992) and play competing roles in determining the fates of amino acids. Cell volume increases also lead to enhancements of amino acid catabolism (Häussinger et al., 1992b), ureagenesis, possibly through increased ASS activity (Quillard et al., 1996), and glutaminase activity (Häussinger et al., 1990). Change in cell hydration, therefore, provides a link for the responses often seen with improved nutrition, i.e. a co-stimulation of protein anabolism and amino acid catabolism.

References Abbott, E.M., Parkins, J.J. and Holmes, P.H. (1985) Influence of dietary protein on the pathophysiology of ovine haemonchosis in Finn Dorset and Scottish Blackface lambs given a single moderate infection. Research Veterinary Science 38, 54–60. Blommart, E.F.C., Luiken, J.J.F.P., Blommart, P.J.E., van Woerkom, G.M. and Meijer, A.J. (1995) Phosphorylation of ribosomal protein S6 is inhibitory for autophagy in isolated rat hepatocytes. Journal of Biological Chemistry 270, 2320–2326. Brosnan, J., Brosnan, M.E., Charron, R. and Nissim, I. (1996) A mass isotopomer study of urea and glutamine synthesis from 15N-labelled ammonia in the perfused rat liver. Journal of Biological Chemistry 271, 16199–16207. Bruckentahl, I., Huntington, G.B., Baer, C.K. and Erdman, R.A. (1997) The effect of abomasal infusion of casein and recombinant somatotropin hormone injection on nitrogen balance and amino acid fluxes in portal-drained viscera and net hepatic and total splanchnic blood in Holstein steers. Journal of Animal Science 75, 1119–1129.

The Liver: Integrator of Nitrogen Metabolism

165

Burrin, D.G., Ferrell, C.L., Eisemann, J.H. and Britton, R.A. (1991) Level of nutrition and splanchic metabolite flux in young lambs. Journal of Animal Science 69, 1082–1091. Burrin, D.G., Britton, R.A., Ferrell, C.L. and Bauer, M.L. (1992) Level of nutrition and visceral organ protein synthetic capacity and nucleic-acid content in sheep. Journal of Animal Science 70, 1137–1145. Calder, P.C. (1995) Glutamine action in cells of the immune system. In: Cynober, L., Fürst, P. and Lawin, P.W. (eds) Pharmacological Nutrition – Immune Nutrition. Zuckschwerdt Verlag, München, pp. 10–20. Calder, A.G., Garden, K.E., Anderson, S.E. and Lobley, G.E. (1999) Quantitation of blood and plasma amino acids using isotope dilution mass spectrometry impact gas chromatography/ mass spectrometry with U-13C amino acids as internal standards. Rapid Communications in Mass Spectrometry 13, 2080–2083. Choung, J.-J. and Chamberlain, D.G. (1995) Effects of intra-ruminal infusion of propionate on the concentrations of ammonia and insulin in peripheral blood of cows receiving intraruminal infusion of urea. Journal of Dairy Research 62, 549–557. Connell, A., Calder, A.G., Anderson, S.E. and Lobley, G.E. (1997) Hepatic protein synthesis in the sheep: effect of intake as monitored by use of stable-isotope-labelled glycine, leucine and phenylalanine. British Journal of Nutrition 77, 255–271. Deuel, T.F., Lovie, M. and Lerner, A. (1978) Glutamine synthetase from rat liver. Journal of Biological Chemistry 253, 6111–6118. Eisemann, J.H. and Huntington, G.B. (1994) Metabolite flux across portal-drained viscera, liver, and hindquarters of hyperinsulinemic, euglycemic beef steers. Journal of Animal Science 72, 2919–2929. Eskild, W., Kindberg, G.M., Smedsrod, B., Blomhoff, R., Norum, K.R. and Berg, T. (1989) Intracellular transport of formaldehyde-treated serum albumin in liver endothelial cells after uptake via scavenger receptors. Biochemical Journal 258, 511–520. Gebhardt, R. (1992) Metabolic zonation of the liver – regulation and implications for liver function. Pharmacological Therapeutics 53, 275–354. Goetsch, A.L. and Ferrell, C.L. (1995) Effects of dietary maize level on net flux across splanchnic tissues of oxygen and nutrients in wethers consuming ad libitum different forages. Animal Science 61, 43–55. Hanigan, M.D., Dijkstra, J., Gerrits, W.J.J. and France, J. (1997) Modelling post-absorptive protein and amino acid metabolism in the ruminant. Proceedings of the Nutritional Society 56, 631–643. Hanigan, M.D., France, J., Wray-Cahen, D., Beever, D.E., Lobley, G.E., Reutzel, L. and Smith, N. W. (1998) Alternative models for analyses of liver and mammary transorgan metabolite extraction data. British Journal of Nutrition 79, 63–78. Häussinger, D. (1990) Nitrogen metabolism in liver: structural and functional organization and physiological relevance. Biochemical Journal 267, 281–290. Häussinger, D. and Gerok, W. (1983) Hepatocyte heterogeneity in glutamate uptake in isolated perfused rat liver. European Journal of Biochemistry 136, 421–425. Häussinger, D., Lang, F., Bauers, K. and Gerok, W. (1990) Interactions between glutamine metabolism and cell volume regulation in perfused rat liver. European Journal of Biochemistry 188, 689–695. Häussinger, D., Lamers, W.H. and Moorman, A.F.M. (1992a) Hepatocyte heterogeneity in the metabolism of amino acids and ammonia. Enzyme 46, 72–93. Häussinger, D., Stoll, B., Morimoto, Y., Lang, F. and Gerok, W. (1992b) Anisoosmotic liver perfusion: redox shifts and modulation of ␣-ketoisocaproate and glycine metabolism. Biological Chemistry Hoppe-Seyler 373, 723–734. Häussinger, D., Lang, F. and Gerok, W. (1994) Regulation of cell function by the cellular hydration state. American Journal of Physiology 267 (Endocrinol. Metab. 30), E343–E355.

166

G.E. Lobley et al. Heitmann, R.N. and Bergman, E.N. (1978) Glutamine metabolism, inter-organ transport and glucogenicity in the sheep. American Journal of Physiology 234, E197–E203. Heitmann, R.N. and Bergman, E.N. (1980) Transport of amino acids in whole blood and plasma of sheep. American Physiological Society 239, E242–E247. Hesketh, J.E., Vasconcelos, M.H. and Bermano, G. (1998) Regulatory signals in messenger RNA: determinants of nutrient–gene interaction and metabolic compartmentation. British Journal of Nutrition 80, 307–321. Houlier, M.L., Mirand, P.P., Durand, D., Bauchart, D., Lefaivre, J. and Bayle, G. (1991) Transport des acides aminés dans l’aire splanchnique par le plasma sanguin et le sang chez le veau préruminant. Reproduction Nutrition Development 31, 399–410. Isserty, A., Ortigues, I. and Remond, D. (1998) Measure of splanchnic blood flow by marker dilution: comparison of 4 methods of para-aminohippuric acid determination. Reproduction Nutrition Development 38, 93–106. Jackson, A.A. (1998) Salvage of urea-nitrogen in the large bowel: functional significance in metabolic control and adaptation. Biochemistry Society Transactions 26, 231–236. Jungermann, K. and Kietzmann, T. (1996) Zonation of parenchymal and nonparenchymal metabolism in liver. Annual Review of Nutrition 16, 179–203. Kaiser, S., Gerok, W. and Häussinger, D. (1988) Ammonia and glutamine metabolism in human liver slices – new aspects on the pathogenesis of hyperammonemia in chronic liver disease. European Journal of Clinical Investigation 18, 535–542. Koeln, L.L., Schlagheck, T.G. and Webb, K.E., Jr (1993) Amino acid flux across the gastrointestinal tract and liver of calves. Journal of Dairy Science 76, 2275–2285. Krehbiel, C.R., Ferrell, C.L. and Freetly, H.C. (1998) Effects of frequency of supplementation on dry matter intake and net portal and hepatic flux of nutrients in mature ewes that consume low-quality forage. Journal of Animal Science 76, 2464–2473. Kuo, F.C., Hwu, W.L., Valle, D. and Darnell, Jr (1991) Co-localisation in pericentral hepatocytes in adult mice and similarity in developmental expression of ornithine aminotransferase and glutamine synthetase mRNA. Proceedings of the National Academy of Sciences USA 88, 9468–9472. Lapierre, H., Reynolds, C.K., Elsasser, T.H., Gaudreau, P., Brazeau, P. and Tyrrell, H.F. (1992) Effects of growth hormone-releasing factor and feed intake on energy metabolism in growing beef steers: Net hormone metabolism by portal-drained viscera and liver. Journal of Animal Science 70, 742–751. Lapierre, H., Bernier, J.F., Dubreuil, P., Reynolds, C.K., Farmer, C., Ouellet, D.R. and Lobley, G.E. (2000) The effect of feed intake level on splanchnic metabolism in growing beef steers. Journal of Animal Science 78, 1084–1099. Lautt, W.W. (1996) Intrinsic regulation of hepatic blood flow. Canadian Journal of Physiology and Pharmacology 74, 223–233. Lescoat, P., Sauvant, D. and Danfaer, A. (1997) Quantitative aspects of protein fractional synthesis rates in ruminants. Reproduction Nutrition Development 37, 493–515. Liu, S.M., Lobley, G.E., MacLeod, N.A., Kyle, D.J., Chen, X.B. and Ørskov, E.R. (1995) Effects of long-term protein excess or deficiency on whole-body protein turnover in sheep nourished by intragastric infusion of nutrients. British Journal of Nutrition 73, 829–839. Lobley, G.E. (1990) Energy metabolism reactions in muscle: responses to age, nutrition and hormonal status. Reproduction Nutrition Development 30, 13–34. Lobley, G.E. (1992) Control of the metabolic fate of amino acids in ruminants: a review. Journal of Animal Science 70, 3264–3275. Lobley, G.E. and Milano, G.D. (1997) Regulation of hepatic nitrogen metabolism in ruminants. Proceedings of the Nutrition Society 56, 547–563. Lobley, G.E., Connell, A., Milne, E., Newman, A.M. and Ewing, T.A. (1994) Protein synthesis in splanchnic tissues of sheep offered two levels of intake. British Journal of Nutrition 71, 3–12. Lobley, G.E. Connell, A., Lomax, M.A., Brown, D.S., Milne, E., Calder, A.G. and Farningham,

The Liver: Integrator of Nitrogen Metabolism

167

D.A.H. (1995) Hepatic detoxification of ammonia in the ovine liver: possible consequences for amino acid catabolism. British Journal of Nutrition 73, 667–685. Lobley, G.E., Connell, A., Revell, D.K., Bequette, B.J., Brown, D.S. and Calder, A.G. (1996a) Splanchnic-bed transfers of amino acids in sheep blood and plasma, as monitored through use of a multiple U-13C-labelled amino acid mixture. British Journal of Nutrition 75, 217–235. Lobley, G.E., Weijs, P.J.M., Connell, A., Calder, A.G., Brown, D.S. and Milne, E. (1996b) The fate of absorbed and exogenous ammonia as influenced by forage or forage-concentrate diets in growing sheep. British Journal of Nutrition 76, 231–248. Lobley, G.E., Bremner, D., Nieto, R., Obitsu, T., Hotston Moore, A. and Brown, D.S. (1998) Transfers of N-metabolites across the ovine liver in response to short-term infusions of an amino acid mixture into the mesenteric vein. British Journal of Nutrition 80, 371–379. Luiken, J.J.F.P., Aerts, J.M.F.G. and Meijer, A.J. (1996) The role of the intralysosomal pH in the control of autophagic proteolytic flux in rat hepatocytes. European Journal of Biochemistry 235, 564–573. Luo, Q.J., Maltby, S.A., Lobley, G.E., Calder, A.G. and Lomax, M.A. (1995) The effect of amino acids on the metabolic fate of 15NH4Cl in isolated sheep hepatocytes. European Journal of Biochemistry 228, 912–917. Lusty, C.J. (1978) Carbamoylphosphate synthetase 1 of rat-liver mitochondria. Purification, properties, and polypeptide molecular weight. European Journal of Biochemistry 85, 373–383. McEvoy, T.G., Robinson, J.J., Aitken, R.P., Findlay, P.A. and Robertson, I.S. (1997) Dietary excesses of urea influence the viability and metabolism of preimplantation sheep embryos and may affect fetal growth among survivors. Animal Reproduction Science 47, 71–90. McLeod, K.R., Bauer, M.L., Harmon, D.L., Reynolds, C.K. and Mitchell, G.E., Jr (1997) Effects of exogenous somatostatin and cysteamine on net nutrient flux across the portaldrained viscera and liver of sheep during intraduodenal infusion of starch hydrolysate and casein. Journal of Animal Science 75, 3026–3037. MacRae, J.C., Bruce, L.A., Hovell, F.D. deB., Inkster, J., Walker, A. and Atkinson, T. (1991) Influence of protein nutrition on the response of growing lambs to exogenous bovine growth hormone. Journal of Endocrinology 130, 53–61. MacRae, J.C., Walker, A., Brown, D. and Lobley, G.E. (1993) Accretion of total protein and individual amino acids by organs and tissues of growing lambs and the ability of nitrogen balance techniques to quantitate protein retention. Animal Production 57, 237–245. Manier, S., Laouari, D., Dechaux, M., Motel, V., Yvert, J.P., Mathan, B. and Kleinknecht, C. (1994) In vivo unaltered muscle protein synthesis in experimental chronic metabolic acidosis. Kidney International 46, 1705–1712. Maxwell, J.L., Terracio, L., Borg, T.K., Baynes, J.W. and Thorpe, S.R. (1990) A fluorescent residualizing label for studies on protein uptake and catabolism in vivo and in vitro. Biochemistry Journal 267, 155–162. May, R.C., Masud, T., Logue, B., Bailey, J. and England, B. (1992) Chronic metabolic acidosis accelerates whole body proteolysis and oxidation in awake rats. Kidney International 41, 1535–1542. Meijer, A.J., Lof, C., Ramos, I.C. and Verhoeven, A.J. (1985) Control of ureagenesis. European Journal of Biochemistry 148, 189–196. Meijer, A.J., Lamers, W.H. and Chamuleau, R.A.F.M. (1990) Nitrogen metabolism and ornithine cycle function. Physiological Reviews 70, 701–748. Milano, G.D. (1997). Liver nitrogen transactions in sheep (Ovis aries). PhD Thesis, University of Aberdeen, UK. Milano, G.D., Lomax, M.A. and Lobley, G.E. (1996) Estimation of the enrichment of urea N precursors. Proceedings of the Nutrition Society 55, 42A. Mortimore, G.E., Poso, A.R. and Lardeux, B.R. (1989) Mechanism and regulation of protein degradation in liver. Diabetes Metabolism Reviews 5, 49–70. Mutsvangwa, T., Buchanan-Smith, J.G. and McBride, B.W. (1997) Effects of ruminally degradable

168

G.E. Lobley et al. nitrogen intake and in vitro addition of ammonia and propionate on the metabolic fate of L-[1-14C]alanine and L-[15N]alanine in isolated sheep hepatocytes. Journal of Animal Science 75, 1149–1159. Nieto, R., Calder, A.G., Anderson, S.E. and Lobley, G.E. (1996) A new method for determination of 15NH3 enrichment in biological samples by gas chromatography/electron ionisation mass spectrometry. Journal of Mass Spectrometry 31, 289–294. O’Sullivan, D., Brosnan, J.T. and Brosnan, M.E. (1998) Hepatic zonation of the catabolism of arginine and ornithine in the perfused rat liver. Biochemical Journal 330, 627–632. Parker, D.S., Lomax, M.A., Seal, C.J. and Wilton, J.C. (1995) Metabolic implications of ammonia production in the ruminant. Proceedings of the Nutrition Society 54, 549–563. Quillard, M., Husson, A. and Lavoinne, A. (1996) Glutamine increases argininosuccinate synthetase mRNA levels in rat hepatocytes. The involvement of cell swelling. European Journal of Biochemistry 236, 56–59. Rattenbury, J.M., Jeacock, M.K. and Shepherd, D.A.L. (1980) Urea synthesis in the liver and kidney of developing sheep. Biochimica et Biophysica Acta 630, 210–219. Reynolds, C.K. (1995) Quantitative aspects of liver metabolism in ruminants. In: Engelhardt, W. von, Leonhard-Marek, S., Breves, G. and Giesecke, D. (eds) Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction: Proceedings of the Eighth International Symposium on Ruminant Physiology. Ferdinand Enke Verlag, Stuttgart, pp. 351–371. Reynolds, C.K. and Tyrrell, H.F. (1991) Effects of mesenteric vein L-alanine infusion on liver metabolism in beef heifers fed on diets differing in forage:concentrate ratio. British Journal of Nutrition 66, 437–450. Reynolds, C.K., Harmon, D.L., Prior, R.L. and Tyrrell, H.F. (1994) Effects of mesenteric vein Lalanine infusion on liver metabolism of organic acids by beef heifers fed diets differing in forage:concentrate ratio. Journal of Animal Science 72, 3196–3206. Rossouw, H.C., van der Walt, J.G. and Nell, M. (1997) Effect of pH and osmolarity on urea metabolism of the liver in small ruminants. Proceedings of the 25th Annual Congress of Physiology Society South Africa, Durban, South Africa, OP47. Rossouw, H.C., Nell, M.J., Mohamed Ali, A. and van der Walt, J.G. (1999) Ammonia partitioning between urea and glutamine in the perfused sheep liver: role of extracellular pH. South African Journal of Animal Science 29, 246–247. Sarraseca, A., Milne, E., Metcalf, M.J. and Lobley, G.E. (1998) Urea recycling in sheep: effects of intake. British Journal of Nutrition 79, 79–88. Stewart, P.M. and Walser, M. (1980) Failure of normal ureagenic response to amino acids in the organic-acid loaded rat: a proposed mechanism for the hyperammonemia of propionic and methylmalonicacidemia. Journal of Clinical Investigation 66, 484–492. Summerskill, W.H.J. and Wolpert, E. (1970) Ammonia metabolism in the gut. American Journal of Clinical Nutrition 23, 633–639. Thorpe, S.R., Baynes, J.W. and Chroneos, Z.C. (1993) The design and application of residualizing labels for studies of protein catabolism. FASEB Journal 7, 399–405. van der Walt, J.G. (1993) Nitrogen metabolism of the ruminant liver. Australian Journal of Agricultural Research 44, 381–403. Webb, K.E., Jr (1999) Tissue, cellular and molecular aspects of peptide absorption and utilization. IX International Symposium on Ruminant Physiology, Pretoria, S. Africa. Welbourne, T.C., Childress, D. and Givens, G. (1986) Renal regulation of interorgan glutamine flow in metabolic acidosis. American Journal of Physiology 226, R858–R866. Whitt, J., Huntington, G., Zetina, E., Casse, E., Taniguchi, K. and Potts, W. (1996) Plasma flow and net nutrient flux across gut and liver of cattle fed twice daily. Journal of Animal Science 74, 2450–2461. Wray-Cahen, D., Roberts, S., Metcalf, J.A., Backwell, F.R.C., Bequette, B.J., Brown, D. S., Sutton, J.D. and Lobley, G.E. (1997) Hepatic response to increased exogenous supply of plasma amino acids by infusion into the mesenteric vein of Holstein-Friesian cows in late gestation. British Journal of Nutrition 78, 913–930.

IV

Tissue Maintenance and Utilization of Endogenous Body Reserves

10

Adipose Tissue: Beyond an Energy Reserve

R.G. VERNON1 AND K.L. HOUSEKNECHT2 1Hannah

Research Institute, Ayr, UK; 2Animal Health Drug Discovery, Pfizer Inc., Groton, Connecticut, USA

Introduction Adipose tissue is a much maligned tissue! The growing prevalence of obesity in the USA and Western Europe in particular has led to a search for means to reduce the amount of adiposity both in ourselves and also in the animals we eat. A number of agents (e.g. growth hormone, ␤-agonists) which can markedly reduce adiposity have been produced and immunological approaches to destroying adipocytes, and hence decreasing adiposity, have been developed. However, in this rush to reduce adiposity, it is worth remembering that adipose tissue has played a key role in mammalian evolution and still has important physiological functions today, both as an energy reserve and as a source of paracrine and endocrine factors.

Role in mammalian evolution and physiology Survival away from an immediate source of food requires a reserve of energy; for terrestrial animals this ideally needs to be light and compact. Triacylgycerol meets these requirements well, being energy dense and also hydrophobic. A gram of stored triacylglycerol in a cell contains virtually no water; in contrast, a gram of stored glycogen contains about 75% water. Triacylglycerol can be stored, in theory, in any cell of the body, but vertebrates evolved a specialist cell type, the adipocyte, for this purpose (Pond, 1992). Poikilotherms store triacylglycerol in skeletal muscle and liver, but in addition have some mesenteric adipose tissue (arguably an extension of the liver as its venous drainage enters the hepatic portal vein); the relatively limited capacity of these species to store energy reflects their much lower energy requirements than homeotherms (Pond, 1992; Sheridan, 1994). The evolution of homeothermy freed animals from some constraints imposed by environmental temperature and allowed exploitation of colder regions, but has a cost as metabolic rate, and hence energy requirements are considerably increased. This increases the need for reserves of energy © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

171

172

R.G. Vernon and K.L. Houseknecht

and is reflected in the much more complex array of adipose tissue depots in mammals and birds than in poikilotherms (Pond, 1992). Detailed anatomical studies by Pond (1986) on a wide variety of mammals has shown that there are some 16 depots, some in the abdominal cavity, some within the musculature and some under the skin, which are present in essentially all mammals including marsupials. The relative size of different depots may vary between species, and in fat animals, depots may effectively merge; such changes are considered to be adaptations to meet the needs of that particular species (Pond, 1992). While adipocytes from different depots appear superficially to be identical, there are subtle differences in their properties, and there is some evidence to suggest that the small depots of the musculature are more metabolically active than the larger abdominal and subcutaneous depots which are usually studied (Pond, 1992). Evolution of adipose tissue has allowed mammals to inhabit some remarkably inhospitable environments; it also facilitates migration, allowing animals to exploit seasonally available food in different locations. Furthermore, adipose tissue has permitted development of a variety of reproductive strategies (Vernon and Pond, 1997). Pregnancy increases the energy requirements of mammals, which are further increased by lactation. Additional demands are mostly met by increasing food intake, but in addition there is often an accumulation of adipose tissue lipid during early pregnancy for use during the later stages of pregnancy, at parturition and during lactation (Vernon and Pond, 1997). Indeed, some species of seal, whale and also boreal bears actually fast during lactation, so deriving all the nutrients required for milk production from tissue reserves (Oftedahl, 1992). Dairy cows, especially those producing large amounts of milk, usually draw heavily on adipose tissue reserves during early lactation, when for reasons as yet unknown, milk output increases more rapidly than appetite. As a result, for a period, mobilization of adipose tissue lipid may account for more than 50% of milk fat production (Vernon and Flint, 1984). Thus adipose tissue has played a central role in mammalian evolution (indeed our evolution would have been impossible without it), and continues to play a key role in mammalian physiology in both wild and domestic species. However, there is a potential cost, at least for wild species, in that too much adipose tissue can reduce mobility and turn an animal into a succulent meal for a predator. Thus the amount of adipose tissue in an animal will usually reflect the relative threats to survival of predation and starvation. For example, African gazelles with a plentiful supply of food have relatively little adipose tissue; in contrast, reindeer in the Arctic have a layer of subcutaneous adipose tissue about 5 cm thick at the beginning of the winter (Larsen et al., 1985). Predators also have this problem as obesity can decrease their chance of obtaining a meal. European foxes are normally thin, but in the Arctic they are markedly fatter (Lindström, 1983). Barnacle geese before migrating from the Arctic to Scotland for the winter, accumulate just enough adipose tissue lipid to meet the needs of the flight (Butler et al., 1998). Animals can thus regulate their levels of adiposity very closely to meet their needs; this means that there must be signals which indicate the size of the body reserves of lipid. Indeed, it is now apparent that adipocytes secrete a variety of substances (Box 10.1) which have autocrine, paracrine and endocrine functions.

Adipose Tissue: Beyond an Energy Reserve

173

Box 10.1. Some substances secreted by adipose tissue cells. Metabolic modulators Lipoprotein lipase Acylation-stimulating protein Apoprotein E Fatty acids Prostaglandin E2 Adenosine Vasoactive factors Prostacyclin (Prostaglandin I2) Monobutyrin Angiotensinogen/Angiotensin II Atrial natriuretic peptide Growth factors IGF-I Transforming growth factors a and ␤ Macrophage colony-stimulating factor

Complement system Factors B, C3 and D (adipsin) Binding proteins IGF-binding proteins Retinol binding protein Cholesterol ester transfer protein Hormones Leptin Oestrone, oestradiol Testosterone Cytokines Tumour necrosis factor a (TNF-a) Interleukin-6 Other Plasminogen activator inhibitor-1 Acrp30/Adipo Q

Factors secreted by adipocytes Adipose tissue secretes a surprisingly large number of substances; they have been roughly categorized in Box 10.1, but several have more than one role (Ailhaud, 1998; Gregoire et al., 1998; Mohammed-Ali et al., 1998). Adipose tissue contains a variety of different types of cell including endothelial cells, undifferentiated adipocyte precursor cells, macrophages, as well as adipocytes. Indeed, in adult animals, adipocytes are a minor component in terms of cell number (10% or less in sheep) (Travers et al., 1997), but being very large cells, adipocytes comprise the bulk of the mass of adipose tissue. While most of the substances listed in Box 10.1 are produced by adipocytes, stromalvascular cells are thought to be the source of steroid hormones (Mohammed-Ali et al., 1998) and interleukin-6 (Fried et al., 1998) produced by adipose tissue. Adipocytes and stromal-vascular cells may also cooperate; thus while adipocytes can produce and release small amounts of prostaglandin E2 and prostacyclin, production is markedly enhanced by the presence of endothelial cells of the stromal-vascular fraction, adipocytes releasing arachidonic acid which is converted to prostaglandin by the endothelial cells (Richelsen, 1992). The fact that a substance is secreted by adipose tissue cells does not necessarily mean that a physiologically significant quantity will enter the general circulation. Arterio-venous difference studies in humans have shown that there is a net release of interleukin-6 but not tumour necrosis factor ␣ (TNF-␣) by subcutaneous adipose tissue (Mohamed-Ali et al., 1997). Arterio-venous difference studies across adipose tissue have also shown a net release of oestradiol and oestrone in women; indeed, adipose

174

R.G. Vernon and K.L. Houseknecht

tissue is the major source of oestrogens in the blood of post-menopausal women (Mohamed-Ali et al., 1998), and adipocytes are the major source of leptin in the general circulation in all states. However, for some substances (e.g. prostaglandins, insulinlike growth factor-I (IGF-I), acylation-stimulating protein), secretion appears to be for autocrine or paracrine effects in adipose or adjacent tissues. For many of the substances listed in Box 10.1, there is no direct evidence for their being produced by adipose tissue in ruminant animals. Prostaglandin E2 has been shown, using microdialysis, to be released by sheep adipose tissue in vivo (Doris et al., 1996). IGF-I, IGF-II (Hovey et al., 1998), IGF-binding protein (Beattie and Vernon, 1995), leptin, as well as lipoprotein lipase have all been shown to be produced by adipose tissue in ruminants.

Autocrine and paracrine effects of locally produced factors within adipose tissue Adipocytes do not divide and are derived from precursor cells within the tissue. The mechanisms and factors which control the rate of proliferation and differentiation of these precursors (which vary with age and adipose tissue depot) are only partially resolved, but include locally produced factors (e.g. IGF-I, prostaglandins), as well as hormones such as glucocorticoids and insulin (Flint and Vernon, 1993; Gregoire et al., 1998; Soret et al., 1999). The production of retinol-binding protein (Tsutsumi et al., 1992) (and also IGF-binding protein) by adipose tissue may also be involved in control of adipogenesis. Retinol itself inhibits adipogenesis (Ohyama et al., 1998), so production of a binding protein may help to attenuate this effect. Interestingly, it seems that retinol may have depot-specific effects on adipogenesis in ruminants (Torii et al., 1996); whether this relates to differences in the ability to produce retinol-binding protein is unknown. Once formed, adipocytes accumulate lipid and can become very large, achieving a size of 2–3 nl in ruminants (Vernon, 1986). However, adipocytes do appear to have a limit to hypertrophic expansion and there is evidence that as they become very large, they induce formation of new adipocytes from precursor cells (Faust et al., 1978). The factors involved in this have not been resolved, but paradoxically, increasing adipocyte size leads to increased production of TNF-␣, which inhibits adipogenesis (Hotamisligil and Spiegelman, 1994). Adipose tissue contains an extensive capillary network (Crandall et al., 1997). As adipocytes enlarge, the ratio of adipocytes to other cells in the tissue diminishes. The identity of the increased number of stromal-vascular cells has not been determined, but it would seem likely that increased numbers of endothelial cells are involved. Adipocytes produce a number of angiogenic factors including monobutyrin and IGF-I (Crandall et al., 1997). The best-known metabolic factor is lipoprotein lipase, which is synthesized and secreted by adipocytes, and then migrates to the luminal surface of endothelial cells where it hydrolyses triacylglycerols of very low density lipoproteins (VLDLs) and chylomicrons (Vernon and Clegg, 1985). Most of the fatty acids so released are taken up by the adipocytes, but some escape and are used elsewhere in the body (Vernon and Clegg, 1985). A second key autocrine factor which promotes fatty acid esterification and glucose uptake in adipocytes is acylation-stimulating protein (Cianflone, 1997).

Adipose Tissue: Beyond an Energy Reserve

175

The curious, chylomicron-dependent synthesis of this protein from a component of the complement system is discussed later (p. 177). Angiotensin II, which is derived from angiotensinogen, also has autocrine effects, increasing lipogenesis in adipocytes (Jones et al., 1997). IGF-I can enhance lipogenesis in adipose tissue in vitro, but this requires a high concentration of the hormone and probably arises from IGF-I interacting with the insulin receptor. As adipocytes (Vernon and Flint, 1989), including sheep adipocytes (S. Wastie, P.J. Buttery and R.G. Vernon, unpublished observations), lack IGF-I receptors this effect is unlikely to be physiologically significant. Obesity results in increased production of TNF-␣ by adipocytes which induces insulin resistance, in part at least due to a decrease in insulin receptor kinase activity (Hotamisligil and Spiegelman, 1994; Gregoire et al., 1998). This effect of TNF-␣ has excited a lot of attention as a possible cause of the non-insulin-dependent diabetes associated with obesity, but the importance of adipocyte derived TNF-␣ in this remains controversial (Schreyer et al., 1998). Indeed, TNF-␣ may not be released from adipose tissue in vivo (Mohammed-Ali et al., 1997), and so may act as an autocrine regulator of adipocyte metabolism. TNF-␣ decreases glucose transport, lipoprotein lipase activity and adipsin production by adipocytes (Hotamisligil and Spiegelman, 1994; Gregoire et al., 1998) and enhances lipolysis. As discussed previously, while animals need a reserve of fat, obesity is usually a disadvantage, so signals which attenuate further fat accumulation as adipocytes enlarge are clearly important. Adipocyte TNF-␣, rather than having a pathological role, may indeed be a key signal of normal physiology for adipocytes. As well as a putative physiological role, production of TNF-␣ and other cytokines may also have pathological roles in some disease states leading to cachexia (Hotamisilgil and Spiegelman, 1994). TNF-␣ production by adipocytes has not been investigated as yet in ruminants, but preliminary studies show that TNF-␣ attenuates insulin stimulation of lipogenesis in sheep adipose tissue in vitro, but is markedly less effective than growth hormone (GH) in this respect (S.E. Melrose, M.D. Houslay and R.G. Vernon, unpublished observations). Leptin, which shares some properties with cytokines such as TNF-␣, may also decrease lipogenesis and enhance lipolysis in rodent adipocytes (Houseknecht and Portocarrero, 1998), but another study found no such effects (Mick et al., 1998). Physiological concentrations of ovine leptin had no effect on either lipogenesis or lipolysis in sheep adipose tissue in vitro (D. Newby, A. Gertler and R.G. Vernon, unpublished observations). Adipose tissue produces prostaglandin E2 and it also produces adenosine (again this is synthesized primarily by stromal-vascular cells by an ectocellular 5¢-nucleotidase) (Vernon et al., 1991). Both substances inhibit catecholamine-stimulated lipolysis (Vernon et al., 1991). Interestingly, catecholamines increase prostaglandin E2 production in vivo and this coincides with a fall in the rate of lipolysis (Doris et al., 1996). Fatty acids released by adipocytes are transported in blood bound to albumin, and albumin supply to adipose tissue can become a limiting factor at high rates of lipolysis (Vernon and Clegg, 1985). Albumin supply is a function of blood flow and blood flow through adipose tissue is higher before a meal (lipolytic state) than immediately after a meal (lipogenic state) (Barnes et al., 1983). Adipose tissue produces monobutyrin, prostacyclin, prostaglandin E2, adenosine, atrial natriuretic peptide and angiotensinogen (converted to angiotensin II within adipose tissue by angiotensinogen-converting enzyme) which are all vasoactive; most are vasodilators but angiotensin is a vasoconstrictor (Crandall et al., 1997). The system is complex, for angiotensin (vasoconstrictor)

176

R.G. Vernon and K.L. Houseknecht

promotes production of prostacyclin (vasodilator) by adipocytes (Crandall et al., 1997). Catecholamines, produced by sympathetic nervous activity in the adipose tissue are also vasoactive (Vernon and Clegg, 1985; Crandall et al., 1997). Indeed, IGF-I may also stimulate blood flow, as there is recent evidence for it stimulating nitric oxide production in endothelial cells, and nitric oxide is again a vasodilator (Böger, 1998). Why so many vasoactive substances are produced by adipose tissue and their relative importance in different physiological situations is unknown. Interestingly, several factors (catecholamines, adenosine, prostaglandin E2) modulate both lipolysis and blood flow in adipose tissue, so control of fatty acid release from adipose tissue in vivo may be more complex than perceived at present. Furthermore, the interaction is not unidirectional, for endothelin-1, which is produced by endothelial cells, inhibits insulin-stimulated glucose uptake in adipocytes (Lee et al., 1998).

Paracrine interactions with neighbouring tissues The widespread but conserved distribution of adipose tissue throughout the mammalian body, and its intimate association with several other tissues, has suggested that adipose tissue may have paracrine effects on these tissues. In young non-pregnant animals, rudimentary mammary glands are found embedded in adipose tissue (Cunha and Hom, 1996; Sheffield, 1988). During pregnancy in rats and mice, the developing mammary tissue spreads through the adjacent inguinal adipose tissue depot and histological examination of mammary tissue during involution reveals a mosaic of clumps of mammary acini and adipocytes. That this association is obligatory rather than coincidental was shown by removing the adipose tissue depot from around the rudimentary mammary gland of young animals and showing that development was severely retarded in the absence of adipose tissue; transplantation of rudimentary mammary glands into adipose tissue resulted in normal development (Cunha and Hom, 1996: Sheffield, 1998). The search for factors involved in this interaction between adipocytes and mammary cells has involved the use of co-culture systems, or culture of mammary tissue with so-called ‘conditioned’ media from adipocyte cultures. Initial studies (Rudland et al., 1984) suggested a role for prostaglandin E2; further studies have suggested an involvement of IGF-I which is both mitogenic and also anti-apoptotic with respect to mammary cells (Hovey et al., 1998; Forsyth et al., 1999). There may well be other factors, and all are not necessarily stimulatory (adipose tissue secretes IGF-binding proteins and transforming growth factor-␤ (TGF␤) which could act to attenuate mammary development) (Rahimi et al., 1998). As well as growth factors, adipocytes may contribute to the extracellular matrix required for mammary development (Wiens et al., 1987). A further aspect is that mammary growth requires angiogenesis, which could also be facilitated by adipocytes. In ruminants, the developing gland does not grow into the adjacent fat pad during pregnancy as in rodents, and involution does not reveal a mosaic of mammary acini and adipocytes (Akers et al., 1990). The mammary gland is, however, associated with adipose tissue and this appears to be critical for prepubertal development (Hovey et al., 1998; Forsyth et al., 1999). The heart has a cap of adipose tissue, the so-called epicardial adipose tissue (Marchington et al., 1989). This depot is present in fetal animals, including fetal rumi-

Adipose Tissue: Beyond an Energy Reserve

177

nants, and the mass of the depot does not correlate with the total mass of adipose tissue in the animal (Marchington et al., 1989). Futhermore, the epicardial adipose tissue abuts directly onto the myocardium without any intervening fascia as is found in skeletal muscle (Marchington et al., 1989). Although circumstantial, these various observations suggest a role for epicardial adipose tissue in the development and functioning of the heart. The former could again relate to the ability of adipose tissue to produce mitogenic factors, while it is possible that cardiac adipocytes provide a direct supply of fatty acids for use by cardiocytes. There is as yet no direct evidence for either possibility and for the latter, there is always a question of whether the rate of supply of fatty acids (presumably transported by diffusion, for there is no obvious portal system in the tissue) would be sufficient for lipolysis in cardiac adipocytes to be a quantitatively significant direct source of fatty acids for heart tissue. The possibilities that inter- and intramuscular adipocytes may influence the development of surrounding skeletal muscle (Hossner et al., 1997), and that they may provide a direct source of fatty acids for adjacent skeletal muscle have also been proposed, but the presence of a fascia between intermuscular adipose tissue and surrounding muscles (Marchington et al., 1989) would suggest that diffusion of fatty acids and other factors is likely to be restricted. The liver receives a direct supply of fatty acids from an adipose tissue depot, for venous drainage from mesenteric adipose tissues leads into the hepatic portal vein. The physiological importance of this is uncertain, but it may have pathological consequences (Barzilai et al., 1999).

Adipose tissue interactions with the immune system Adipose tissue interacts with the immune system in several ways: by production of cytokines, proteins of the complement system and leptin and by interactions between adipocytes and cells of the lymph nodes. Complement acts to lyse, and hence destroy, targeted cells during infection. Adipocytes secrete three components of the alternative system for complement production, factors B, C3 and D; factor D is also known as adipsin (Flier, 1995; Cianflone, 1997). Adipsin production is decreased by insulin and TNF-␣ and is decreased during obesity in rodents but apparently not in obese humans (Flier, 1995). The physiological significance of this adipocyte-based system for complement production is still unclear. Curiously, factors C3 and B bind to chylomicrons and this promotes cleavage of factor C3 by adipsin to produce factor C3A which, in turn, is cleaved by carboxypeptidase to produce factor C3A desarg (acylation-stimulating protein) (Cianflone, 1997). This raises the possibility that the primary function of factor B, C3 and D production by adipocytes is autocrine and concerned with adipocyte metabolism, rather than paracrine modulation of complement formation. Lymph nodes are embedded in adipose tissue, mostly in minor depots that are too small to make a significant contribution to whole body lipid storage and which are selectively conserved in starvation (Pond, 1996). Studies with an intermuscular (popliteal) adipose tissue depot have shown that both the lipolytic response and the fatty acid composition of triacylglycerols of adipocytes varies with distance from the lymph nodes in guinea pigs (Pond, 1996). It appears that lymphocytes secrete cytokines

178

R.G. Vernon and K.L. Houseknecht

(e.g. TNF-a) which accentuate lipolysis (Feingold et al., 1992; Green et al., 1994), while the fatty acids released from adipocytes stimulate lymphocyte proliferation (Pond, 1996). Stimulation of the immune system by injection of an endotoxin (lipopolysaccharide) results in enhanced lipolysis and the appearance of TNF-␣ type I receptors on adipocytes close to the lymph node but not on adipocytes remote from the lymph node (MacQueen and Pond, 1998; Pond and Mattacks, 1998). However, studies with ovine popliteal adipose tissue failed to detect any variation in fatty acid composition with distance from the lymph node (Vernon and Pond, 1997), and TNF␣ had no effect on basal or catecholamine-stimulated lipolysis in adipose tissue near to (5 mm or less) or distant from the lymph node (S.E. Melrose, M.D. Houslay and R.G. Vernon, unpublished observations). However, preliminary studies suggest that the ability of the adenosine analogue, N6-phenylisopropyladenosine, to inhibit lipolysis appears to be diminished in adipocytes close to the lymph nodes, and culture with growth hormone diminished the ability of PIA to inhibit lipolysis in adipocytes distant from, but not in those close to, the lymph node (S.E. Melrose, M.D. Houslay and R.G. Vernon, unpublished observations). Thus there may be differences in the lipolytic system in different parts of popliteal adipose tissue in sheep, but, if so, they appear to be much smaller than in guinea pigs. Leptin also influences the immune system. Impaired T-lymphocyte immunity is found in ob/ob mice which produce a defective form of leptin (Chandra, 1980), and very recent studies show that leptin enhances proliferation of T-lymphocytes and their subsequent ability to respond to antigens (Lord et al., 1998). In addition, preadipocytes may also contribute to the body’s defence mechanisms as there is evidence for their having macrophage-like properties (Cousin et al., 1999); this may explain the production of macrophage colony-stimulatory factor by adipose tissue and its promotion of adipose tissue growth (Levine et al., 1998). Thus there are several ways in which adipose tissue can modulate functioning of the immune system, but their relative importance is uncertain. That such interactions are physiologically relevant is indicated by the fact that anorexia, which results in greatly diminished adiposity, causes diminished functioning of the immune system (Chandra, 1991). High-yielding dairy cows tend to be more susceptible to disease (e.g. mastitis, laminitis) and this is exacerbated by poor nutrition (Sinclair et al., 1999); adiposity is likely to markedly diminish in such animals. From an individual’s point of view, for starvation, and other conditions which lead to a loss of adipose tissue reserves, to result in a failure of the immune system seems singly unfortunate. However, when considered from the perspective of the species as a whole, it does make sense, for when food is in short supply it increases the likelihood that the weakest animals, with the least reserves, die leaving more of the limited food available for the stronger. A case of the survival of the fattest!

Leptin The discovery of leptin provided the long-postulated link between adipose tissue stores and brain centres which regulate food intake and energy homeostasis. Since this seminal discovery, our understanding of the role of leptin has expanded from that of a ‘sensor’ of body fat mass to include participation in, and regulation of, multiple physiological systems including reproduction, inflammation and cell-mediated immunity, as

Adipose Tissue: Beyond an Energy Reserve

179

well as the coordination of whole body energy homeostasis (Friedman and Halaas, 1998; Houseknecht and Portocarrero, 1998; Houseknecht et al., 1998a). The leptin (ob) gene has been cloned in several species including the mouse, pig, chicken and human (Houseknecht and Portocarrero, 1998) and also the cow (Ji et al., 1998); sequence homology among species is high. Leptin is primarily expressed in white adipocytes; however, in rodents and humans, leptin expression has also been reported in the placenta (Hoggard et al., 1998) and stomach (Bado et al., 1998). In cattle and swine, leptin is reported to be expressed exclusively in adipose tissue (Houseknecht and Portocarrero, 1998). Furthermore, Ji et al. (1998) reported no difference in the level of leptin gene expression among adipose tissue depots (subcutaneous, renal and omental depots) in cattle, which differs from reports in humans (Montague et al., 1998; Russell et al., 1998; Vanharmelen et al., 1998). The leptin receptor is a member of the cytokine family of receptors and is expressed as either a long form, or as multiple short forms due to alternative splicing (Tartaglia, 1997). The long form of the leptin receptor predominates in the hypothalamus and is credited with the central regulation of appetite; the short isoforms, and to a limited extent, the long form, are expressed in multiple peripheral tissues as well and may be involved in the regulation of tissue metabolism (Tartaglia, 1997). Soluble forms of the leptin receptor have been reported (Houseknecht et al., 1996); their role is uncertain, but leptin binding to proteins in blood complicates the accurate assay of circulating leptin in many species. Dyer et al. (1997a) cloned a partial ovine long-form leptin receptor cDNA and reported that mRNA expression for the long-form receptor is present in the hypothalamus, anterior pituitary and adipose tissue of sheep. Furthermore, they reported that expression was highest in the arcuate nucleus and the ventromedial hypothalamus, and that hypothalamic expression was up-regulated by undernutrition as found in rodents (Houseknecht and Portocarrero, 1998). Leptin is a powerful regulator of food intake in rodents; effects are most dramatic in ob/ob mice which lack leptin (Friedman and Halaas, 1998; Houseknecht et al., 1998a; Houseknecht and Portocarrero, 1998). Intracerebro ventricular administration of recombinant ovine leptin to sheep also caused a reduction in food intake (Morrison et al., 1998). Large-scale preparation of recombinant ruminant leptin is now underway (Gertler et al., 1998) which will facilitate further studies in ruminants. Studies on the mechanism whereby leptin decreases appetite and increases whole-body energy expenditure have, in the past, focused mostly on the neuropeptide Y (NPY) system of the hypothalamus, but other neuropeptides are most probably involved. Leptin acts centrally to prevent NPY-induced appetite stimulation and suppression of thermogenesis by inhibiting its synthesis in the arcuate nucleus of the hypothalamus (Friedman and Halaas, 1998; Houseknecht et al., 1998a). Dyer et al. (1997b) showed that peripheral NPY administration up-regulated the expression of leptin and NPY receptor subtype 1 in ovine adipose tissue. These data are indicative of a feedback loop between leptin and NPY expression which has been shown for rodents by others (Houseknecht et al., 1998a; Houseknecht and Portocarrero, 1998). Leptin expression is under complex control by both hormones (Friedman and Halaas, 1998; Houseknecht et al., 1998a) and the sympathetic nervous system (Trayhurn et al., 1998); modulatory hormones include insulin, glucocorticoids, thyroid hormones and oestrogen. Effects appear to be chronic and due to changes in leptin gene expression rather than to acute stimulation of leptin secretion from intracellular

180

R.G. Vernon and K.L. Houseknecht

storage pools. In vitro studies from cattle (Houseknecht et al., 1998b) show that leptin mRNA abundance in bovine subcutaneous adipose tissue is increased by insulin and glucocorticoids. In vivo, growth hormone increased leptin gene expression in bovine (Houseknecht et al., 1998b) and ovine adipose tissue (Raymond et al., 1997). However, effects of GH may be indirect, as leptin expression in bovine adipose tissue was strongly, positively correlated with IGF-I gene expression while GH had no effect on leptin expression in vitro (Houseknecht et al., 1998b). Serum leptin concentration in the fed state is highly, positively correlated with body fat mass in rodents, pigs and humans (Friedman and Halaas, 1998; Houseknecht et al., 1998a) and in cattle (Ji et al., 1997; Chilliard et al., 1998; Minton et al., 1998) and sheep (Kumar et al., 1998). This led to the hypothesis that leptin allows the body to ‘sense’ the size of the body’s energy storage pool and adjust appetite and whole-body energy metabolism accordingly. Fasting causes a rapid, profound down-regulation of leptin gene expression in rodents and humans (Friedman and Halaas, 1998). Fasting (2 days) also down-regulates leptin gene expression in subcutaneous adipose tissue of cattle (Tsuchiya et al., 1998) and swine (Spurlock et al., 1998a). However, the effects of fasting on leptin expression in pigs are modest compared with rodents, and there appears to be no effect of maintenance intake on leptin gene expression (Spurlock et al., 1998a), suggesting that this species may be less sensitive to this form of regulation of the leptin gene. Fasting also results in changes in the secretion of pituitary hormones, including thyroid stimulating hormone (TSH), GH and the gonadotropins, and these changes are partly prevented by administration of leptin (Friedman and Halaas, 1998). These various observations indicate that leptin plays a central role in the adaptations to fasting and suggest that the primary evolutionary role for leptin is to prevent death by starvation rather than to prevent obesity (Spiegelman and Flier, 1996). Leptin is an important regulator of reproduction in rodents and humans, providing a link between energy reserves and an energetically demanding function (Hoggard et al., 1998; Houseknecht et al., 1998a; Macut et al., 1998). There is potential for leptin to act both centrally and peripherally in both males and females, as the leptin receptor is expressed not only in the hypothalamus and pituitary but in the ovary and testis as well (Houseknecht et al., 1998a). Leptin stimulates gonadotrophin secretion in ob/ob mice and undernourished animals, and advances the onset of puberty (Houseknecht et al., 1998a). These data have obvious, important implications for productive efficiency of livestock. Links between leptin and the immune system have been discussed in a previous section. In addition, leptin expression is up-regulated by inflammatory cytokines such as TNF-␣ and interleukins, and an increase in leptin secretion may cause the anorexia following administration of endotoxin (Houseknecht et al., 1998a). However, Spurlock et al. (1998b) found no effect of acute endotoxin treatment on leptin expression in fasted pigs, despite physiological indications of a robust inflammatory response. Additionally, Leininger et al. (1998) found that endotoxin treatment down-regulated leptin gene expression in fully fed pigs, and that the endotoxin-induced fall in leptin was highly correlated with changes in blood insulin, IGF-I, glucose and free fatty acid concentrations. These data (Leininger et al., 1998) suggest that changes in energy metabolism and associated hormones can overcome, at least in the pig, the positive effects of inflammatory cytokines on leptin gene expression. Obviously, further work is

Adipose Tissue: Beyond an Energy Reserve

181

necessary to elucidate the role of leptin (if any) in the anorexia and altered energy metabolism associated with acute and chronic inflammation in livestock.

Conclusions Ideas on adipose tissue have progressed a long way from the days when it was perceived as a relatively inert tissue. Not only do we now know that it is very active metabolically, it is also a source of a plethora of factors which influence events within adipose tissue, within adjacent tissues or indeed with endocrine effects on distant parts of the body – a veritable ‘node on the information superhighway’ (Flier, 1995). The importance of adipose tissue in vertebrate evolution, and mammalian evolution in particular, is now apparent, and its central role in mammalian physiology is beginning to be appreciated. This is emphasized by two very recent studies with transgenic mice essentially devoid of adipose tissue; many mice died before reaching adulthood and those that survived showed symptoms of acute diabetes (McKnight 1998); the reason for the diabetes is uncertain but may be due to an ability to remove fatty acids from the blood. Curiously, both too little and too much adipose tissue can lead to the same pathological condition! Animals have affected means of adjusting their amount of adipose tissue to meet their needs, although the mechanisms may be attenuated by domestication. Some of the substances produced by adipocytes and their associated stromal-vascular cells must act as signals to help achieve this. Leptin, initially perceived as the factor which could solve the problem of obesity, may, in fact, be more important for signalling inadequate or declining adipose tissue reserves; leptin clearly does not prevent the development of

Insulin

(–) Adipsin

ASP

(–) TNF-a (+)

TRIACYLGLYCEROL

Precursor supply

Leptin (–) Appetite (+) Energy expenditure

Fig. 10.1. Mechanisms whereby leptin and TNF-a attenuate triacylglycerol accumulation. ASP, acylation-stimulating protein; (+) stimulates, (-) inhibits, ≠ increased, Ø decreased.

182

R.G. Vernon and K.L. Houseknecht

obesity, but it may retard it. Leptin also has an important role in the adaptations to negative energy balance. As adipocytes enlarge, they increase production of other substances (e.g. TNF-␣) as well as leptin, and decrease production of others (e.g. adipsin, acylation-stimulation protein) which may act to limit further expansion. Adipose tissue mass thus appears to be modulated both directly by autocrine factors and indirectly by endocrine factors such as leptin (Fig. 10.1). When one chooses to ignore these signals, obesity follows and thence undesirable pathological conditions (e.g. diabetes). The implication of this for domestic ruminants (and indeed other food animals) is that we need to know more precisely the amounts of adipose tissue required for optimal function, for while too much adipose tissue is undesirable both for the animal and as a food for us, equally too little is deleterious for the animal’s well-being.

References Ailhaud, G. (1998) L’adipocyte, cellule sécrétrice et endocrine. Médecine/sciences 14, 858–864. Akers, R.M., Beal, W.E., McFadden, T.B. and Capuco, A.V. (1990) Morphometric analysis of involuting bovine mammary tissue after 21 or 42 days of non-suckling. Journal of Animal Science 68, 3604–3613. Bado, A., Levasseur, S., Attoub, S., Kermorgant, S., Laigneau, J.P., Bortoluzzi, M.N., Moizo, L., Lehy, T., Guerre-Millo, M., Marchand-Brustel, Y. le and Lewin, M.J.M. (1998) The stomach is a source of leptin. Nature 394, 790–793. Barnes, R.J., Comline R.S. and Dobson, A. (1983) Changes in the blood flow to the digestive organs of sheep induced by feeding. Quarterly Journal of Experimental Physiology 68, 77–88. Barzilai, N., She, L., Liu, B.-Q., Vuguin, P., Cohen, P., Wang, J. and Rossetti, L. (1999) Surgical removal of visceral fat reverses hepatic insulin resistance. Diabetes 48, 94–98. Beattie, J. and Vernon, R.G. (1995) Glucocorticoids regulate the secretion of a 21kDa-IGFbinding protein by sheep adipose tissue explants. Molecular Cellular Biochemistry 145, 151–157. Böger, R.H. (1998) Role of nitric oxide in the haemodynamic effects of growth hormone. Growth Hormone and IGF Research 8, 163–165. Butler, P.J., Woakes, A.J. and Bishop, C.M. (1998) Behaviour and physiology of Svalbard Barnacle Geese Branta leucopsis during their autumn migration. Journal of Avian Biology 29, 536–545. Chandra, R.K. (1980) Cell-mediated immunity in genetically obese (C57BL/6J ob/ob) mice. American Journal of Clinical Nutrition 33, 13–16. Chandra, R.K. (1991) Nutrition and immunity: lessons from the past and new insights into the future. American Journal of Clinical Nutrition 53, 1087–1101. Chilliard, Y., Ferlay, A., Delavaud, C. and Bocquier, F. (1998) Plasma leptin in underfed or overfed adult holstein and charolais cows and its relationship with adipose tissue cellularity. International Journal of Obesity 22 (Suppl. 3), S171. Cianflone, K. (1997) Acylation stimulating protein and the adipocyte. Journal of Endocrinology 155, 203–206. Cousin, B., Munoz, O., Andre, M., Fontanilles, A.M., Dani, C., Cousin, J.L., Laharrague, P., Casteilla, L. and Penicaud, L. (1999) A role for preadipocytes as macrophage-like cells. FASEB Journal 13, 305–312. Crandall, D.L., Hausman, G.J. and Kral, J.G. (1997) A review of the microcirculation of adipose tissue: anatomic, metabolic and angiogenic perspectives. Microcirculation 4, 211–232. Cunha, G.R. and Hom, Y.K. (1996) Role of mesenchymal-epithelial interactions in mammary gland development. Journal of Mammary Gland Biology and Neoplasia 1, 21–35.

Adipose Tissue: Beyond an Energy Reserve

183

Doris, R., Thompson, G.E., Finley, E., Kilgour, E., Houslay, M.D. and Vernon, R.G. (1996) Chronic effects of somatotropin treatment on response of subcutaneous adipose tissue lipolysis to acutely acting factors in vivo and in vitro. Journal of Animal Science 74, 562–568. Dyer, C.J., Simmons, J.M., Matteri, R.L. and Keisler, D.H. (1997a) Leptin receptor mRNA is expressed in ewe anterior pituitary and adipose tissues and is differentially expressed in hypothalamic regions of well-fed and feed-restricted ewes. Domestic Animal Endocrinology 14, 119–128. Dyer, C.J., Simmons, J.M., Matteri, R.L. and Keisler, D.H. (1997b) Effects of an intravenous injection of NPY on leptin and NPY-Y1 receptor mRNA expression in ovine adipose tissue. Domestic Animal Endocrinology 14, 325–333. Faust, I.M., Johnson, P.R., Stern, J.S. and Hirsh, J. (1978) Diet-induced adipocyte number increase in adult rats: a new model of obesity. American Journal of Physiology 235, E279E286. Feingold, K.R., Doerrler, W., Dinarello, C.A., Fiers, W. and Grunfeld, C. (1992) Stimulation of lipolysis in cultured fat cells by tumor necrosis factor, interleukin-1 and the interferons is blocked by inhibition of prostaglandin synthesis. Endocrinology 130, 10–16. Flier, J.S. (1995) The adipocyte: storage depot or node on the energy information superhighway? Cell 80, 15–18. Flint, D.J. and Vernon, R.G. (1993) Hormones and adipose growth. In: Schreibman, M.P., Scanes, C.G. and Pang, P.K.T. (eds) The Endocrinology of Growth, Development and Metabolism in Vertebrates. Academic Press, San Diego, pp. 469–499. Forsyth, I.A., Gabai, G. and Morgan, G. (1999) Spatial and temporal expression of insulin-like growth factor-I, insulin-like growth factor-II and the insulin-like growth factor-I receptor in the sheep fetal mammary gland. Journal of Dairy Research 66, 35–44. Fried, S.K., Bunkin, D.A. and Greenberg, A.S. (1998) Omental and subcutaneous adipose tissues of obese subjects release interleukin-6: depot difference and regulation by glucocorticoid. Journal of Clinical Endocrinology and Metabolism 83, 847–850. Friedman, J.M. and Halaas, J.L. (1998) Leptin and the regulation of body weight in mammals. Nature 395, 763–770. Gertler, A., Simmons, J. and Keisler, D.H. (1998) Large-scale preparation of biologically active recombinant ovine obese protein (leptin). FEBS Letters 422, 137–140. Gregoire, F.M., Smas, C.M. and Sul, H.S. (1998) Understanding adipocyte differentiation. Physiological Reviews 78, 783–809. Green, A., Dobias, S.B., Walters, D.J.A. and Brasier, A.R. (1994) Tumor necrosis factor increases the rate of lipolysis in primary cultures of adipocytes without altering levels of hormonesensitive lipase. Endocrinology 134, 2581–2588. Hoggard, N., Hunter, L., Trayhurn, P., Williams, L.M. and Mercer, J.G. (1998) Leptin and reproduction. Proceedings of the Nutrition Society 57, 421–427. Hossner, K.L., Yemm, R., Vierck, J. and Dodson, M.V. (1997) Insulin-like growth factor (IGF)-I and -II and IGFBP secretion by ovine satellite cell strains grown alone or in coculture with 3T3-L1 preadipocytes. In Vitro Cellular and Developmental Biology 33, 791–795. Hotamisligil, G.S. and Spiegelman, B.M. (1994) Tumor necrosis factor a: a key component of the obesity-diabetes link. Diabetes 43, 1271–1278. Houseknecht, K.L. and Portocarrero, C.P. (1998) Leptin and its receptors: Regulators of wholebody energy homeostasis. Domestic Animal Endocrinology 15, 457–475. Houseknecht, K.L., Mantxoros, C.S., Kuliawat, R., Hadro, E.T., Flier, J.S. and Kahn, B.B. (1996) Evidence for leptin binding to proteins in serum of rodents and humans: Modulation with obesity. Diabetes 45, 1638–1643. Houseknecht, K.L., Baile, C.A., Matteri, R.L. and Spurlock, M.E. (1998a) The biology of leptin: A review. Journal of Animal Science 76, 1405–1420. Houseknecht, K.L., Portocarrero, C.P., Ji, S., Lemenager, R.P. and Spurlock, M.E. (1998b).

184

R.G. Vernon and K.L. Houseknecht Growth hormone (GH) regulation of leptin gene expression in bovine adipose tissue: In vitro and in vivo studies. International Journal of Obesity 22 (Suppl. 3), S166 (abstract). Hovey, R.C., Davey, H.W., Mackenzie, D.D.S. and McFadden, T.B. (1998) Ontogeny and epithelial-stromal interactions regulate IGF expression in the ovine mammary gland. Molecular and Cellular Endocrinology 136, 139–144. Ji, S.Q., Willis, G.M., Scott, R.R. and Spurlock, M.E. (1997) Partial cloning of the bovine leptin gene and its expression in adipose depots and in cattle before and after finishing. Journal of Animal Science 75 (Suppl. 1), 167. Ji, S.Q., Willis, G.M., Scott, R.R. and Spurlock, M.E. (1998) Partial cloning and expression of the bovine leptin gene. Animal Biotechnology 9, 1–14. Jones, B.H., Standridge, M.K. and Moustaid, N. (1997) Angiotensin II increases lipogenesis in 3T3-L1 and human adipose cells. Endocrinology 138, 1512–1519. Kumar, B., Francis, S.M., Suttie, J.M. and Thompson, M.P. (1998) Expression of obese mRNA in genetically lean and fat selection lines of sheep. Comparative Biochemistry and Physiology 120, 543–548. Larsen, T.S., Nilsson, N.Ö. and Blix, A.S. (1985) Seasonal changes in lipogenesis and lipolysis in isolated adipocytes from Svalbard and Norwegian reindeer. Acta Physiological Scandinavia 123, 97–104. Lee, Y.-C., Juan, C.-C., Fang, V.S., Hsu, Y.-P., Lin, S.-H., Kwok, C.-F. and Ho, L.-T. (1998) Evidence that endothelin-1 (ET-1) inhibits insulin-stimulated glucose uptake in rat adipocytes mainly through ETA receptors. Metabolism 47, 1468–1471. Leininger, M.T., Portocarrero, C.P., Bidwell, C.A., Spurlock, M.E., Nielsen, J.N. and Houseknecht, K.L. (1998) Response to immune challenge in pigs selected for high lean gain: Role of leptin. Journal of Animal Science 76 (Suppl. 1), 129. Levine, J.A., Jensen, M.D., Eberhardt, N.L. and O’Brien, T. (1998) Adipocyte macrophage colony-stimulating factor is a mediator of adipose tissue growth. Journal of Clinical Investigation 101, 1557–1564. Lindström, E. (1983) Condition and growth of red foxes (Vulpes vulpes) in relation to food supply. Journal of Zoology, London 199, 117–122. Lord, G.M., Matarese, G., Howard, J.K., Baker, R.J., Bloom, S.R. and Lechler, R.I. (1998) Leptin modulates the T-cell immune response and reverses starvation-induced immunosuppression. Nature 394, 897–900. McKnight, S.L. (1998) WAT-free mice: diabetes without obesity. Genes and Development 12, 3145–3148. MacQueen, H.A. and Pond, C.M. (1998) Immunofluorescent localisation of tumour necrosis factor-a receptors on the popliteal lymph node and the surrounding adipose tissue following a simulated immune challenge. Journal of Anatomy 192, 223–231. Macut, D., Micic, D., Pralong, F.P., Bischof, P. and Campana, A. (1998) Is there a role for leptin in human reproduction? Gynaecology Endocrinology 12, 321–326. Marchington, J.M., Mattacks, C.A. and Pond, C.M. (1989) Adipose tissue in the mammalian heart and pericardium: structure, foetal development and biochemical properties. Comparative Biochemistry and Physiology 94B, 225–232. Mick, G., Vanderbloomer, T., Fu, C.L. and McCormick, K. (1998) Leptin does not affect adipocyte glucose metabolism: studies in fresh and cultured adipocytes. Metabolism 47, 1360–1365. Minton, J.E., Bindel, D.J., Drouillard, J.S., Titgemeyer, E.C., Grieger, D.M. and Hill, C.M. (1998) Serum leptin is associated with carcass traits in finishing cattle. Journal of Animal Science 76 (Suppl. 1), 231. Mohamed-Ali, V., Goodrick, S., Rawesh, A., Katz, D.R., Miles, J.M., Yudkin, J.S., Klein, S. and Coppack, S.W. (1997) Subcutaneous adipose tissue releases interleukin-6, but not tumor necrosis factor-a, in vivo. Journal of Clinical Endocrinology and Metabolism 82, 4196–4200.

Adipose Tissue: Beyond an Energy Reserve

185

Mohamed-Ali, V., Pinkney, J.H. and Coppack, S.W. (1998) Adipose tissue as an endocrine and paracrine organ. International Journal of Obesity 22, 1145–1158. Montague, C.T., Prins, J.B., Sanders, L., Zhang, J.L., Sewter, C.P., Digby, J., Byrne, C.D. and O’Rahilly, S. (1998). Depot-related gene expression in human subcutaneous and omental adipocytes. Diabetes 47, 1384–1391. Morrison, C.D., Daniel, J.A., Holmberg, B.J., Bolden, O.U., Raver, N., Gertler, A. and Keisler, D.H. (1998) Effects of lateral cerebroventricular infusion of leptin on ewe lambs. Journal of Animal Science 76 (Suppl. 1), 225. Oftedal, O.T. (1992) The adaptation of milk secretion to the constraints of fasting in bears, seals and baleen whales. Journal of Dairy Science 76, 3234–3246. Ohyama, M., Matsuda, K., Torii, S., Matsui, T., Yano, H., Kawada, T. and Ishihara, T. (1998) The interaction between vitamin A and thiazolidinedione on bovine adipocyte differentiation in primary culture. Journal of Animal Science 76, 61–65. Pond, C.M. (1986) The natural history of adipocytes. Science Progress Oxford 70, 45–71. Pond, C.M. (1992) An evolutionary and functional view of mammalian adipose tissue. Proceedings of the Nutrition Society 51, 367–377. Pond, C.M. (1996) Interactions between adipose tissue and the immune system. Proceedings of the Nutrition Society 55, 111–126. Pond, C.M. and Mattacks, C.A. (1998) In vivo evidence for the involvement of the adipose tissue surrounding lymph nodes in immune responses. Immunology Letters 63, 159–167. Rahimi, N., Tremblay, E., McAdam, L., Roberts, A. and Elliott, B. (1998) Autocrine secretion of TGF-ß1 and TGF-ß2 by preadipocytes and adipocytes: a potent negative regulator of adipocyte differentiation and proliferation of mammary carcinoma cells. In Vitro Cellular and Developmental Biology 34, 412–420. Raymond, S.R., Thomas, M.G., Carroll, J.A., Matteri, R.L. and Keisler, D.H. (1997) Zeranol and growth hormone treatment differentially influenced mRNA levels of the obesity protein, leptin, and the GH receptor in growing wethers. Journal of Animal Science 75 (Suppl. 1), 225. Richelsen, B. (1992) Release and effects of prostaglandins in adipose tissue. Prostaglandins Leukotrienes and Essential Fatty Acids 47, 171–182. Rudland, P.S., Davies, A.C.T. and Tsao, S.-W. (1984) Rat mammary preadipocytes in culture produce a trophic agent for mammary epithelia – prostaglandin E2. Journal of Cellular Physiology 120, 364–376. Russell, C.D., Petersen, R.N., Rao, S.P., Ricci, S.P., Prasad, A., Zhang, Y., Brolin, R.E. and Fried, S.K. (1998) Leptin expression in adipose tissue from obese humans: depot-specific regulation by insulin and dexamethasone. American Journal of Physiology 38, E507-E515. Schreyer, S.A., Chua, S.C. and LeBoeuf, R.C. (1998) Obesity and diabetes in TNF-␣ receptordeficient mice. Journal of Clinical Investigation 102, 402–411. Sheffield, L.G. (1988) Organisation and growth of mammary epithelia in the mammary gland fat pad. Journal of Dairy Science 71, 2855–2874. Sheridan, M.A. (1994) Regulation of lipid metabolism in poikilothermic vertebrates. Comparative Biochemistry and Physiology 107B, 495–508. Sinclair, M.C., Nielsen, B.L., Oldham, J.D. and Reid, H.W. (1999) Consequences for immune function of metabolic adaptations to load. British Society of Animal Science Occasional Publication 24, 113–118. Soret, B., Lee, H.-J., Finley, E., Lee, S.C. and Vernon, R.G. (1999) Regulation of differentiation of sheep subcutaneous and abdominal preadipocytes in culture. Journal of Endocrinology 161, 517–524. Spiegelman, B.M. and Flier, J.S. (1996) Adipogenesis and obesity-rounding out the big picture. Cell 87, 377–389. Spurlock, M.E., Frank, G.R., Cornelius, S.G., Ji, S.Q., Willis, G.M. and Bidwell, C.A. (1998a) Obese gene expression in porcine adipose tissue is reduced by food deprivation but not by maintenance or submaintenance intake. Journal of Nutrition 128, 677–682.

186

R.G. Vernon and K.L. Houseknecht Spurlock, M.E., Ranalletta, M.A., Cornelius, S.G., Frank, G.R., Willis, G.M., Ji, S., Grant, A.L. and Bidwell, C.A. (1998b) Leptin expression in porcine adipose tissue is not increased by endotoxin but is reduced by growth hormone. Journal of Interferon Cytokine Research 18, 1051–1058. Tartaglia, L.A. (1997) The leptin receptor. Journal of Biological Chemistry 272, 6093–6096. Torii, S., Matsui, T. and Yano, H. (1996) Development of intramuscular fat in Wagyu beef cattle depends on adipogenic or antiadipogenic substances present in serum. Animal Science 63, 73–78. Travers, M.T., Vernon, R.G. and Barber, M.C. (1997) Repression of the acetyl-coA carboxylase gene in ovine adipose tissue during lactation: the role of insulin responsiveness. Journal of Molecular Endocrinology 19, 99–107. Trayhurn, P., Duncan, J.S., Hoggard, N. and Rayner, D.V. (1998) Regulation of leptin production: a dominant role for the sympathetic nervous system? Proceedings of the Nutrition Society 57, 413–419. Tsuchiya, T., Nagao, Y., Ozawa, A., Matsumoto, M., Sugahara, K., Kubo, T. and Kato, H. (1998) Decrease of the obese gene expression in bovine subcutaneous adipose tissue by fasting. Bioscience Biotechnology and Biochemistry 62, 2068–2069. Tsutsumi, C., Okuno, M., Tannous, L., Piantedosi, R., Allan, M., Goodman, D.S. and Blaner, W.S. (1992) Retinoids and retinoid-binding protein expression in rat adipocytes. Journal of Biological Chemistry 267, 1805–1810. Vanharmelen, V., Reynisdottir, S., Eriksson, P., Thorne, A., Hoffstedt, J., Lonnqvist, F. and Arner, P. (1998) Leptin secretion from subcutaneous and visceral adipose tissue in women. Diabetes 47, 913–917. Vernon, R.G. (1986) The growth and metabolism of adipocytes. In: Buttery, P.J., Haynes, N.B. and Lindsay, D.B. (eds) Control and Manipulation of Animal Growth. Butterworths, London, pp. 67–83. Vernon, R.G. and Clegg, R.A. (1985) The metabolism of white adipose tissue in vivo and in vitro. In: Cryer, A. and Van, R.L.R. (eds) New Perspectives in Adipose Tissue. Butterworths, London, pp. 65–86. Vernon, R.G. and Flint, D.J. (1984) Adipose tissue: metabolic adaptation during lactation. Symposium of the Zoological Society of London 51, 119–145. Vernon, R.G. and Flint, D.J. (1989) Role of growth hormone in the regulation of adipocyte growth and function. In: Heap, R.G., Prosser, C.G. and Lamming, G.E. (eds.) Biotechnology in Growth Regulation. Butterworths, London, pp. 57–71. Vernon, R.G. and Pond, C.M. (1997) Adaptations of maternal adipose tissue to lactation. Journal of Mammary Gland Biology and Neoplasia 2, 231–241. Vernon, R.G., Finley, E. and Watt, P.W. (1991) Adenosine and the control of adrenergic regulation of adipose tissue lipolysis during lactation. Journal of Dairy Science 74, 695–705. Wiens, D., Park, C.S. and Stockdale, F.E. (1987) Milk protein expression and ductal morphogenesis in the mammary gland in vitro: hormone-dependent and -independent phases of adipocyte-mammary epithelial cell interaction. Developmental Biology 120, 245–258.

11

Regulation of Growth and Metabolism During Postnatal Development

B.H. BREIER, M.H. OLIVER AND B.W. GALLAHER Research Centre for Developmental Medicine and Biology, Faculty of Medicine and Health Science, University of Auckland, Auckland, New Zealand

Introduction This chapter will review the endocrine regulation of growth and metabolism in ruminants and focus on recent insights into the central role of the somatotropic axis. Growth and metabolism during postnatal life are regulated by multiple hormones and growth factors acting in an endocrine (systemic) and an autocrine or paracrine (local) manner. The somatotropic axis is a multi-level hormonal system, primarily consisting of growth hormone (GH; somatotropin) and insulin-like growth factor-I (IGF-I; somatomedin), their associated carrier proteins and receptors. There are a number of other hormonal axes that are involved in the regulation of intermediary metabolism and the growth process. Among these are insulin, thyroid hormones, glucocorticoids, sex steroids, the melanocortin-leptin axis and a number of locally produced growth factors. Circulating GH is a major regulator of metabolism and growth during postnatal life. GH deficiency is associated with a reduction in growth rate and in final adult size. During postnatal development and during adult life circulating GH regulates key metabolic pathways of intermediary metabolism (Breier et al., 1991). Treatment with GH is lipolytic, elevating circulating concentrations of free fatty acids (FFA), increasing the body’s ability to respond to lipolytic stimuli and reducing lipogenesis. GH treatment also creates a state of positive nitrogen balance by increasing nitrogen retention and decreasing protein catabolism. Furthermore, GH administration elevates plasma insulin and glucose levels, which in the long term creates a state of insulin resistance. IGF-I was originally described as an endocrine factor secreted from the liver in response to GH stimulation which mediated the effects of GH in peripheral tissues. However, while IGF-I treatment increases growth in normal and in GH-deficient animals, growth rates rarely equal those obtained with GH treatment. GH stimulation of IGF-I mRNA expression is not limited to the liver, suggesting that some of the effects of GH treatment on tissue growth are derived from the direct stimulation of autocrine or paracrine IGF-I production. Treatment with IGF-I stimulates protein metabolism © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

187

B.H. Breier et al.

188

and reduces plasma glucose, as a result of both increased glucose uptake and reduced glucose production (Douglas et al., 1991). IGF-I is important for the function of a variety of organs. For example, elevated circulating IGF-I levels increase glomerular filtration rate (Hammerman and Miller, 1993) and renal growth, while local concentrations of IGF-I determine the destiny of maturing follicles in the ovary (Giudice et al., 1995). IGF-I is also a potent inotropic agent in the heart and regulates immune cell numbers as well as other aspects of the immune response. While IGF-I is often considered a proliferation or differentiation factor, it is also a potent cell survival factor through its role as an inhibitor of programmed cell death. This review will elucidate the complexity of the regulation of growth and metabolism during postnatal life by the somatotropic axis using four main examples: developmental changes, effects of nutrition, GH and IGF-I treatment, and altered GH action in conditions of metabolic imbalance.

Regulation of postnatal growth by the somatotropic axis Circulating growth hormone Growth hormones and lactogenic hormones form a family of biochemically related hormones; they are single-chain polypeptides with two or three intrachain disulphide bridges and about 200 residues. There is considerable variation between species in the structure of GH, prolactin (PRL) and placental lactogen (PL), which in some cases can be correlated with significant variation in biological activity. Regulation of GH synthesis and secretion from the pituitary primarily reflects interactions between GH releasing factor (GRF) and inhibition by somatostatin (SRIF). GH secretion is initiated firstly by a reduction in the secretion of SRIF and an increase in the release of GRF from the hypothalamus into the hypophyseal portal system. Subsequently the pituitary enters a period of GRF-reduced responsiveness which generates the characteristic pulsatile pattern of GH secretion. The pattern of GH secretion is sexually dimorphic, with males exhibiting large pulses and low basal levels while females exhibit small pulses and high baseline levels (Gatford et al., 1998). The secretion of GH is also regulated by a series of feedback mechanisms. For example, GH directly feeds back to inhibit its own secretion, while IGF-I inhibits GH gene transcription and the stimulation of GH release from the pituitary by GRF. Circulating GH is a major regulator of postnatal growth. Subjects with acromegaly (excessive pituitary GH secretion) exhibit bone overgrowth in the form of increased height and enlarged hands and feet resulting from the stimulatory effects of GH on osteoblast activity and bone mineral deposition. GH deficiency is associated with a reduction in growth rate and in final adult size. In the adult animal circulating GH regulates diverse metabolic pathways including protein, fat and carbohydrate metabolism.

Growth hormone receptors The GH receptor (GHR) is a member of the class 1 cytokine receptor family. GHR mRNA and immunoreactivity have been detected in most tissues postnatally, with highest abundance in liver and adipose tissue (Bingham et al., 1994). The GH molecule

Regulation of Growth and Metabolism During Postnatal Development

189

has two binding sites for the extracellular domain of the GHR and forms a 1:2 GH:GHR complex by sequential homodimerization. Ilondo et al. (1994) demonstrated in adipocytes that the receptor dimerization process is critical for GH-stimulated glucose incorporation into lipid. GH binding to the GHR activates a number of intracellular pathways. Transcription of the IGF-I gene is rapidly increased soon after GH administration, suggesting that it is an immediate consequence of GHR signalling. The GHR may also interact with cell membrane associated G proteins to regulate the metabolic effects of GH (Roupas and Herington, 1994). The multiple physiological and metabolic actions of GH could partly be explained by heterogeneity of the GHR. Heterogeneity of the GHR has been demonstrated by ligand binding studies, by affinity cross-linking, by epitope mapping with monoclonal antibodies and by molecular studies. The soluble growth hormone binding protein (GHBP) can also be viewed as a variant form of the GHR. The GHBP represents the extracellular domain of the GHR, although its derivation appears to be species-specific. In rodents, a separate alternatively spliced mRNA exists for GHBP, conferring on the protein a unique hydrophilic 17-residue. In some species plasma GHBP arises from proteolysis of the full-length GHR protein (Baumann, 1994), while in humans, sheep and chickens the GHBP is thought to be mainly produced by proteolytic cleavage of the GHR extracellular domain (Bingham et al., 1994). The GHBP in ruminant plasma has been investigated by Davis et al. (1992) using a charcoal separation assay, demonstrating the presence of GHBP in cattle and sheep. Devolder et al. (1993) provided further evidence for the existence of GHBP in cattle using size-exclusion chromatography and ligand and Western blotting techniques. Three major specific bands that bind GH were demonstrated; two of these bands (190 and 58 kDa) were recognized by a monoclonal antibody directed against a part of the extracellular domain of the bovine growth hormone receptor (bGHR). The function of GHBP is not clear at present, however, it is thought to prolong the half-life of GH in circulation and may play a role in tissuespecific delivery. Since the GHBP is regulated by a number of physiological variables, it is possible that the GHBP may represent an additional mechanism that regulates GH action.

Insulin-like growth factors The IGFs are a family of polypeptides related structurally and evolutionarily to proinsulin. The classical somatomedin hypothesis, which postulated that the effects of GH are mediated by circulating IGFs of hepatic origin (endocrine action), has been modified. It is now generally accepted that IGFs are synthesized at multiple sites and the concept of local action of GH in many tissues to promote both cell differentiation and proliferation through local production of IGF-I acting in an autocrine or paracrine manner has become firmly established (Isaksson et al., 1991). Although feedback mechanisms between the two systems exist, the relative importance of each mechanism is uncertain. The significance of endocrine IGF-I is supported by the observation that the selection of mice (Blair et al., 1988) or sheep (Sherlock et al., 1993) on the basis of high or low plasma concentrations of IGF-I results in a marked divergence of their growth rate. While IGF-I treatment increases growth in the GH-deficient state, growth rates rarely equal those obtained with GH, which may reflect that some of the effects of

B.H. Breier et al.

190

GH treatment on tissue growth are derived from the direct stimulation of autocrine or paracrine IGF-I production. As will be discussed later, the endocrine and the paracrine or autocrine mode of action are interrelated and contribute to a coordinated adaptive process. An interesting concept in the regulation of IGF-I action relates to the IGF-I gene itself. The IGF-I gene consists of at least six exons dispersed over a region of more than 80 kb (Rotwein, 1991). Exons 3 and 4 encode the mature IGF-I peptide. The synthesis of IGF-I is a complex process involving two promoters, multiple sites of transcription initiation, alternate splicing and multiple polyadenylation sites (Adamo et al., 1993). The observation that exons 1 and 2 are differentially spliced to exon 3, producing alternate class 1 and class 2 transcripts, has stimulated discussion about endocrine versus paracrine or autocrine functions of the IGF-I peptide. Changes in abundance of hepatic class 2 transcripts have been associated with concomitant changes in circulating (endocrine) IGF-I. For example, Pell et al. (1993) suggest that in sheep, hepatic IGF-I output may be regulated by nutrition and GH, primarily through class 2 transcripts. Thus, in conditions of optimal growth, the increase in circulating IGF-I correlates with an increased abundance of exon 2 transcripts in liver IGF-I mRNA. However, in peripheral tissues like muscle, exon 1 IGF-I transcripts predominate and do not respond markedly to a range of metabolic and endocrine stimuli (Gilmour, 1994). The switching of class 1 and class 2 transcripts in different tissues, depending on the endocrine and metabolic conditions of the animal, could represent an adaptive process to accommodate the changing needs of different tissues. While the growth promoting properties of IGF-I are well established, the function of IGF-II in ruminants is not clear, although there is some evidence that IGF-II can impair the anabolic effects of IGF-I in sheep (Koea et al., 1992a). It is known that at the cellular level, IGF-II has proliferative actions, enhancing proliferation of ovarian granulosa cells and stimulating wound repair processes such as keratinocyte proliferation. A proliferative role for IGF-II is also evident in many cancerous tissues (Isaksson et al., 1991).

IGF binding proteins and receptors The IGF binding proteins (IGFBPs) are a family of peptides which exhibit high affinity for the IGFs. Six high affinity IGFBPs (IGFBP-1 to -6) have been identified in blood plasma to date. There is strong sequence homology at their N and C termini, especially with respect to the presence of 16 Cys residues, indicating the likely importance of these regions in IGF binding. There is considerably more sequence heterogeneity between the intermediate portions of IGFBPs suggesting that this region may define many of the variations in activity between the binding proteins. The IGFBPs vary considerably in size, their affinity for IGFs and other molecules, phosphorylation and glycosylation states, and protease sensitivity. The regulation and function is specific to each of the IGFBPs. Each of the IGFBPs could have a unique role as a modulator of IGF action and the different IGFBPs appear to be under independent endocrine and metabolic regulation. Changes in the production of IGFBPs may initiate tissue-specific targeting and distribution of IGFs. In addition, either systemic or locally produced

Regulation of Growth and Metabolism During Postnatal Development

191

IGFBPs may alter the distribution of circulating IGFs within target tissues by modifying receptor–IGF interactions, thus enhancing or reducing biological activity. IGFBP-1 The circulating IGFBP-1 peptide is approximately 26 kDa in size and has a similar affinity for IGF-I and IGF-II. IGFBP-1 is under strong metabolic control and is produced mainly by hepatocytes, although some expression is also found in the kidney and uterine decidua. Serum levels of IGFBP-1 are high in the fetus but decline postnatally. Plasma levels of IGFBP-1 are inversely proportional to those of insulin (Lee et al., 1993) and rapidly increase in response to treatment with stress-related hormones such as glucocorticoids, glucagon and catecholamines (Hooper et al., 1994). The regulation of IGF-I availability by IGFBP-1 plays an acute glucose counter-regulatory function, restricting the insulin-like actions of the IGFs during states of a relative insulin deficiency (Baxter, 1995). IGFBP-2 The IGFBP-2 protein is approximately 32 kDa in size and has two distinct but overlapping IGF binding sites, of which the IGF-II site exhibits a fivefold greater affinity for its ligand than the IGF-I site. Circulating IGFBP-2 is mainly produced by hepatocytes (Scharf et al., 1995) but it is also present at high concentrations in milk and seminal plasma. Plasma levels of IGFBP-2 in the postnatal circulation vary in response to many factors including ontogeny, nutritional status and GH status. Despite a striking inverse relationship between plasma concentrations of GH and IGFBP-2, the regulation of IGFBP-2 is unlikely to occur directly via GH action (Gallaher et al., 1995). The role of plasma IGFBP-2 in the functional regulation of circulating IGFs is unclear. IGFBP-2 levels are high in catabolic states where IGFBP-3 is suppressed and it may restrict IGFI availability in such situations. However, since the sensitivity of IGFBP-2 gene transcription to insulin is much lower compared with that of IGFBP-1, IGFBP-2 is more likely be a chronic, as opposed to an acute, regulator of IGF action. IGFBP-3 IGFBP-3 is expressed in a wide range of tissues and undergoes extensive post-translational modifications. IGFBP-3 is secreted in a phosphorylated state but the effect on IGF affinity is minimal. However, since charged regions on IGFBP-3 are important in ternary complex formation in plasma and cell surface binding, phosphorylation has been implicated in serum half-life or tissue distribution (Coverly and Baxter, 1995). The predominant IGF carrier in the postnatal circulation is a 150 kDa ternary complex consisting of IGF-I or IGF-II, IGFBP-3 and an acid labile subunit (ALS). Circulating IGFBP-3 is predominantly derived from the non-parenchymal cells of the liver (Chin et al., 1994) while ALS and endocrine IGF-I are produced by adjacent hepatocytes. The ternary complex is formed firstly by the association of IGF-I or -II with IGFBP-3, followed by the binding of ALS to the IGF–IGFBP-3 dimer. When IGFs are incorporated in the ternary complex they are considerably more stable than the free peptide. The binary complex IGFBP-3/IGF-I has a much shorter circulating half-life and rapidly crosses into the extravascular compartment where it is targeted to specific tissues (Arany et al., 1993). Plasma IGFBP-3 correlates with GH and IGF-I, and is a chronic indicator of GH-dependent growth status. However, this relationship is indirect and reflects the GH dependency of ALS and IGF-I production.

B.H. Breier et al.

192

IGFBP-4, -5 and -6 The mature IGFBP-4 peptide is 24 kDa in size and is n-glycosylated to create a 28 kDa form. The function of IGFBP-4 glycosylation is unclear although one role may be in the regulation of vascular endothelial transport. Unlike other IGFBPs, IGFBP-4 persistently inhibits IGF action. For example, IGFBP-4 is elevated and IGF-I activity is suppressed in atretic ovarian follicles but the opposite is observed in dominant follicles (Giudice et al., 1995). The IGFBP-5 gene is expressed in a tissue-specific fashion and the mature peptide is approximately 29 kDa in size, with phosphorylation and glycosylation variants (Jones and Clemmons, 1995). IGFBP-5 generally enhances the effects of IGFs on cell proliferation. IGFBP-6 mRNA is widely expressed and yields a mature peptide that is approximately 22 kDa in size. It has a strong preference for binding IGF-II over IGF-I. IGFBP-6 has been found in plasma, amniotic fluid, follicular fluid and cerebrospinal fluid (CSF) in sheep and humans. Levels in serum decline slightly from mid to late gestation but are stable after birth (Lewitt et al., 1995). IGF receptors The classical IGF-I receptor (IGF-IR) is structurally homologous to the insulin receptor, forming a heterotetramer. The ␣ subunit, 90 kDa in size, is involved in ligand binding and forms the extracellular portion of the receptor. The affinity of the IGF-IR is highest for IGF-I, twofold lower for IGF-II and 100-fold reduced for insulin. Following the binding of IGF-I to the IGF-IR, there is an activation of the receptor tyrosine kinase activity which results in the phosphorylation of insulin receptor substrate I (IRS-I). The mitogenic actions of IGF-I appear to be exclusively mediated through the IGF-IR. In cells where both IGF-I and insulin receptor subunits are both synthesized, hybrid receptors (IGF-HR) can form by the association of an (␣␤) subunit from each of the IGF-IR and insulin receptors to create a heterotetramer. IGF-I is bound to these receptors with an affinity similar to that of the IGF-IR, and IGF-II binds with fourfold lower affinity. Insulin binds with a very low affinity to the IGF-HR relative to IGF-I and also relative to insulin binding to its own receptor (Soos et al., 1994). While the physiological significance of the IGF-HR is unclear, the sequestration of IR subunits into less insulin-responsive hybrids might be a mechanism to modulate insulin sensitivity in tissues. The IGF-II receptor (IGF-II/M6PR) exclusively binds IGF-II with high affinity and is identical to the cation-independent mannose-6-phosphate receptor, consisting of a short intracellular/membrane spanning domain and a large extracellular domain made up of multiple structural repeats (Kiess et al., 1994). In rat tissues, IGF-II/M6PR immunoreactivity and mRNA abundance is high in the fetus but very low after birth, suggesting a minimal role in postnatal development and an active role in fetal growth and/or development. Also deduced from IGF gene knockout experiments (Baker et al., 1993) is the existence of a further IGF-II receptor (IGF-XR) which is expected to mediate a number of the effects of IGF-II on fetal and placental development.

Developmental changes within the somatotropic axis There are crucial differences in the mechanisms that regulate growth and development before and after birth. The regulation of fetal growth depends on maternal phenotype

Regulation of Growth and Metabolism During Postnatal Development

193

and the stage of gestation. Fetal growth in late gestation is limited by maternal constraint. The primary elements ensuring adequate fetal growth are the ability of the mother to provide adequate nutritional support to the placenta and the fetus, and the ability of the fetus to manage the maternal resources appropriately. One of the key roles of endocrine factors during fetal growth and development therefore is provided by hormones which influence the partitioning of nutrients between mother, placenta and fetus, and which regulate the ability of the fetus to utilize the available substrate. Postnatal growth is predominantly determined by genotype and by nutrition. Animals of different genotype grow within predictable boundaries and the growth process is under central endocrine control of pituitary GH.

Perinatal transition of the somatotropic axis from a fetal to a postnatal state Fetal growth and development is rapidly adapted to conditions of maternal and/or placental substrate limitation. After birth the neonate does not need this acute link to the mother and the developmental switch to a mature somatotropic axis allows growth to be directly linked to environmental factors. In sheep, normal parturition is triggered by a rise in fetal cortisol which begins approximately 10–15 days before term and increases markedly in the last 3–5 days before delivery (Fowden et al., 1996). The essential role of the rise in fetal glucocorticoids on lung maturation are well known. However, the maturational effects of the glucocorticoids may have a far more general role in switching the fetal physiological state to a postnatal state. Glucocorticoids play a major role in the induction of hepatic gluconeogenic enzymes in the perinatal period, since gluconeogenesis is a major metabolic pathway for neonatal adaptation. Glucocorticoids also have important effects on cardiac development and a variety of endocrine systems. Major changes occur during the immediate prepartum and neonatal period at many levels of the somatotropic axis. Plasma concentrations of GH are reduced gradually during the last weeks before birth (Breier and Gluckman, 1991) and there is a major decrease in GH secretion at the time when the final cortisol surge is observed prior to parturition. The expression of GHR mRNA and GHR receptor number are very low during mid- or late gestation (Klempt et al., 1993), particularly in the liver, until after birth. As described above, there is a marked increase in plasma IGF-I and a marked decline in plasma IGF-II after birth, and the respective hepatic mRNA levels are also altered accordingly (Li et al., 1993). Plasma concentrations of ovine IGFBP-1 and IGFBP-2 and the circulating M6P/IGF-IIR during the perinatal and neonatal period markedly decrease while IGFBP-3 concentrations increase markedly after birth (Gallaher et al., 1994). Similarly, IGF-IR concentrations in skeletal muscle decline markedly after birth. The mechanisms of these changes within the somatotropic axis are not well understood at present. However, the perinatal increase in hepatic GHR and the rise in IGF-I are not activated if the cortisol rise is abolished by lesioning of the paraventricular nucleus (Breier et al., 1994) or by adrenalectomy (Li et al., 1996). These studies suggest that the prepartum cortisol surge plays an important role in initiating the perinatal switch of the somatotropic axis from the fetal to the postnatal state and function generally observed during postnatal life.

B.H. Breier et al.

194

Postnatal ontogeny of the somatotrophic axis The ontogeny of GH secretion in the perinatal period has been most extensively studied in the sheep. In the fetal lamb, GH concentrations are markedly elevated compared with postnatal concentrations (Breier and Gluckman, 1991). Within 24 h of birth, plasma GH concentrations fall tenfold with both a reduction in pulse amplitude and interpulse values to a secretory pattern not markedly different from the adult. However, sexually differentiated patterns of GH secretion are documented in a number of species, with a higher pulse amplitude and lower baseline values in the male than in the female. Recent work by Gatford et al. (1998) suggests that GH secretion in ram lambs shows markedly higher GH pulse amplitudes than in ewe lambs, and that this sexually dimorphic pattern of GH secretion may be regulated by differences in portal GRF secretion. It is well established that gonadal steroids and anabolic compounds enhance GH release in ruminants through a direct stimulatory effect on GH synthesis and release at the pituitary level. The increase in baseline concentrations of GH in oestradioltreated steers independent of nutritional plane (Breier et al., 1988a) is compatible with such a conclusion. In most species there is a gradual increase in hepatic GHR expression during the first weeks or months after birth (Breier et al., 1988b). This rise in hepatic GHR levels correlates with an increase in plasma concentrations of IGF-I and with the postnatal growth curve. The coordinated increase in GHR concentrations and plasma IGF-I suggests that GHR activity may be a sensitive indicator of increased tissue responsiveness to GH after birth (Breier et al., 1994). A recent study by Min et al. (1999) investigated the endocrine responsiveness of sheep at different ages, ranging from 7 days after birth until one year of age. While the number of GHRs in the liver did not differ in the basal state, GH treatment increased the hepatic GHR number only after puberty. Similarly, the IGF-I response to GH was markedly higher in yearling sheep in comparison with prepubertal animals (Min et al., 1999). In ruminants, as in other species, plasma levels of IGF-I are low at birth and show a postnatal rise which is concomitant with the major increase of GHR in the liver (Breier et al., 1994). A pubertal rise of plasma IGF-I is well documented and this rise is related to sex steroid production. However, in lambs sex differences in plasma IGF-I already exist before weaning with ram lambs showing twofold higher IGF-I levels than ewe lambs (Gatford et al., 1998) and this sexual dimorphism is maintained after puberty. The abundance of IGF-II mRNAs is high before birth and declines postnatally although this developmental pattern of regulation is subject to tissue specificity (Delhanty and Han, 1993). An ontogenic reduction is seen in fetal plasma levels of IGF-II. However, plasma IGF-II still remains higher than IGF-I postnatally in sheep. While IGF-II is integral to adequate prenatal growth in the mouse (Baker et al., 1993), researchers have struggled to show a correlation between IGF-II and birth weight, and its role in postnatal growth is not clearly defined, being stimulatory in some studies but not in others. The ability of IGF-II to affect glucose clearance is significantly less than IGF-I (Douglas et al., 1991) and unlike IGF-I, IGF-II does not ameliorate protein catabolism in lambs, but rather acts as an IGF-I antagonist. Although there is some evidence that plasma IGFBP-1 concentrations are high in the fetus and decline postnatally (Liu et al., 1991), there is little data on IGFBP-1 in

Regulation of Growth and Metabolism During Postnatal Development

195

ruminants. Given that IGFBP-1 is under strong metabolic control and that plasma levels of IGFBP-1 are inversely proportional to those of insulin, there is considerable scope for further research in this area. IGFBP-2 is the most abundant IGFBP prior to birth and, although plasma concentrations fall after birth, it is the second most abundant carrier protein in adult sheep plasma (Gallaher et al., 1995). Recent work in sheep confirms data from other species which shows that plasma concentrations IGFBP-2 are high during fetal life, they fall gradually after birth and plateau after one year of age. Levels of IGFBP-3 observed in fetal sheep plasma are significantly lower than those in the adult (Gallaher et al., 1998) and during postnatal life there is a distinct ontogenic increase in plasma concentrations of IGFBP-3 in sheep, similar to the ontogenic pattern observed for IGF-I (Gatford et al., 1998). The marked ontogenic changes in IGFIR concentrations serve as additional evidence for a key role of IGF-I during fetal development. There is a rapid decline of IGF-IR levels in skeletal muscle; the concentrations in fetal muscle are about tenfold higher than in postnatal muscle (Boge et al., 1995).

Influence of nutrition on the somatotropic axis It is now widely accepted that nutritional status plays a major role in determining circulating GH concentrations. Plasma levels of GH are elevated as a result of undernutrition in ruminants, reflecting the influence of such factors as hypoglycaemia, stress and low serum free fatty acids on the pituitary secretion of GH (Breier and Gluckman, 1991). In sheep and cattle the GH response to GRF is enhanced during restricted feeding compared with ad libitum feeding. This suggests an explanation for the increase in GH secretion during reduced nutritional intakes, which is related to diminished negative feedback within the somatotropic axis mediated in part via a decrease of hypothalamic SRIF (Thomas et al., 1991). Undernutrition is also associated with reduced hepatic GH binding capacity in lambs (Sauerwein et al., 1991) and steers (Breier et al., 1988c) leading to a state of relative GH resistance which may reflect an inhibition of GHR dimerization or altered signal transduction. Plasma concentrations of IGF-I are clearly dependent on nutritional status. The ability of GH to increase plasma IGF-I is impaired under conditions of reduced nutritional intake when the number of GHRs is reduced (Breier et al., 1988a). Basal concentrations of IGF-I and the response to GH are reduced during periods of insufficient protein and/or energy intake. Dietary protein supply seems to be the limiting factor for a maximal stimulation of IGF-I plasma concentrations by GH (Elsasser et al., 1989). The relative GH resistance in poorly fed animals and consequently the inverse changes in plasma GH and IGF-I concentrations may represent a mechanism for preferential utilization of mobilized substrates to maintain homeostasis rather than cell growth and proliferation. While transcription of all hepatic IGF-I mRNA variants is reduced by undernutrition in the sheep, the effect is greatest on the GH sensitive (endocrine IGF-I) exon 2 transcripts (Pell et al., 1993). In the rat, fasting is associated with a widespread reduction in hepatic and most extra-hepatic IGF-I mRNA expression, although brain and heart expression is minimally affected (Olchovsky et al., 1993). Preservation of IGF-I synthesis in these tissues may represent the mechanism by which critical tissues and

B.H. Breier et al.

196

organs are protected during catabolic states. Conversely, IGF-II is slightly increased following undernutrition in adult sheep (Hua et al., 1995). This observation together with the demonstration that co-administration of IGF-II blocks the reduction in protein catabolism induced by IGF-I treatment in the fasted lamb (Koea et al., 1992b) invites the hypothesis that the increase in the plasma IGF-II/IGF-I ratio during undernutrition may exacerbate the development of a catabolic state. The effect of undernutrition on the IGFBPs is varied. Consistent with its glucose counter-regulatory function, IGFBP-1 gene transcription and serum concentrations are significantly and rapidly elevated by reduced food intake. Nutritional restriction generally results in an increase in plasma IGFBP-2 which occurs via an increase in transcription rate. This increase has been specifically linked to a reduction in protein intake (Smith et al., 1995). However, there is still some controversy about the regulation of IGFBP-1 and IGFBP-2 in farm animals. Fasting of newborn pigs results in an increase in plasma levels of IGFBP-1 but a decrease in IGFBP-2 levels (McCusker et al., 1991). In response to 72 h starvation in adult sheep, IGFBP-1 is not significantly changed, while IGFBP-2 increases (Gallaher et al., 1992). A clear inverse relationship has been demonstrated recently in sheep between plasma insulin and glucose concentrations and plasma concentrations of IGFBP-1 and IGFBP-2, suggesting that both are inversely regulated by plasma insulin (Gallaher et al., 1995). While the changes in plasma IGFBP-1 and -2 following undernutrition appear to restrict the insulin-like activities of the IGFs during catabolic states, the changes in IGFBP-3 and -4 may reflect an attempt to maximize the availability of the remaining IGFs to the tissues. Circulating levels of IGFBP-3 and the ternary complex (IGF storage capacity) are reduced (Gallaher et al., 1992) while specific IGFBP-3 protease activity to reduce IGF affinity for IGFBP-3 is enhanced. In conditions of reduced nutritional intake, IGFBP4, an inhibitor of IGF action, is decreased both at the serum and transcriptional level (Holt et al., 1996).

Effects of GH and IGF-I on protein and energy metabolism The endocrine and metabolic effects of GH treatment are well established. Exogenous GH treatment induces re-partitioning of nutrients leading to increased lipolysis, increased protein synthesis, decreased protein degradation and increased bone growth. While it is established that GH has a dominant influence on IGF-I synthesis, the precise mode of action of GH treatment is complex. In the lamb, chronic treatment with bGH led to a dose-dependent increase in the number of high-affinity hepatic GHRs, plasma concentrations of IGF-I and growth rate, while carcass protein and fat content showed positive and negative dose-response relationships respectively (Sauerwein et al., 1991). Body weight gain and protein content of the carcass showed significant positive correlations with plasma IGF-I. Treatment with GH can even improve protein metabolism during an acute catabolic state (simple 72 h starvation) (Ogawa et al., 1996). Beginning GH therapy before or during the onset of starvation has comparable effects on the amelioration of protein breakdown. A number of components of the somatotropic axis are influenced by GH treatment, which contribute to the coordination of nutrient partitioning between different tissues. The up-regulation of the GHR and stimulation of post-receptor mechanisms may lead to increased synthesis of IGF-I and

Regulation of Growth and Metabolism During Postnatal Development

197

a rise in its plasma concentrations. Since plasma IGFBP-3 is increased while IGFBP-2 is decreased with GH treatment, this change in relative concentrations of IGFBPs may facilitate an increased availability of IGF-I at target tissues. Treatment with bGH of lambs significantly reduces plasma concentrations of IGFBP-2 while nutritional restriction increases plasma concentrations of IGFBP-2. Since these changes observed in plasma concentrations of IGFBP-2 show a negative correlation with changes in plasma IGF-I, it has been suggested that IGF-I itself may be an important regulating factor for the availability of IGFs (Gallaher et al., 1995). There is increasing evidence for ontogenic changes in responsiveness to GH treatment at identical doses. The treatment with GH (bGH at 0.3 mg kg⫺1 day⫺1) of wellfed healthy adult ewes leads to a marked reduction in protein breakdown and elevated circulating concentrations of IGF-I. In addition, a marked diabetogenic effect on plasma glucose and insulin is observed, and a major rise in plasma FFA levels shows significant lipolytic action in the healthy adult state (Hennies et al., 1998). While the lipolytic action of GH is well recognized, it is minimal in young animals or in conditions of enhanced metabolic demand such as lactation, where high levels of energy expenditure and utilization prevent a major rise in plasma FFA. There is no increase in plasma FFA in young pre-pubertal lambs with GH (bGH at 0.3 mg kg⫺1 day⫺1). In addition, these lambs showed a comparatively modest increase in plasma IGF-I and a very mild diabetogenic response with only slightly elevated plasma glucose and insulin levels during the well-fed phase of the study (Ogawa et al., 1996). This comparison suggests that the same dose of GH leads to somewhat different actions depending on the age of the animal. Consistent with its growth promoting activity, IGF-I administration stimulates protein metabolism and it reduces protein catabolism in the lamb, while passive immunization against IGF-I in lambs elevates net protein catabolism (Koea et al., 1992b). Treatment with IGF-I also has insulin-like effects on glucose metabolism. IGF-I reduces plasma glucose as a result of both increased glucose uptake and reduced glucose production. However, in the sheep such data reflect a pharmacological response since at physiological doses, IGF-I suppresses plasma insulin but increases glucose concentrations. Only as the dose increases does IGF-I saturate the IGFBPs, ‘free IGF-I’ levels are increased and a direct effect on glucose metabolism is observed (Douglas et al., 1991). There is general agreement that elevation of plasma concentrations of IGF-I by treatment with high doses of IGF-I shows characteristic insulin-like effects inducing hypoglycaemia. This effect depends on saturation of the IGFBPs in circulation leading to a marked increase in ‘free IGF-I’ which binds to the insulin receptor. Two different experimental paradigms were developed to investigate the metabolic effects of short-term IGF-I infusion (5 h infusion) in the 48-h fasted lamb and longterm effects of 8-hourly IGF-I injections for 8 weeks in well-fed lambs. In short-term treatment studies, young lambs were infused with IGF-I for 5 h after a 48-h period of starvation (Douglas et al., 1991). Two doses of IGF-I were used: a low dose which caused no elevation of ‘free IGF-I’ in the circulation, and a higher dose which caused a sevenfold elevation of ‘free IGF-I’. The lower dose did not alter plasma concentrations of glucose, or the rate of glucose production or clearance, while the higher dose caused mild hypoglycaemia. When this higher dose of IGF-I and a dose of insulin of equivalent hypoglycaemic potential were compared, only IGF-I increased protein synthesis in skeletal muscle, heart and liver. Both IGF-I infusions at the low and at the high dose

198

B.H. Breier et al.

markedly reduced protein degradation and this effect may be mediated via the IGF-I receptor (Douglas et al., 1991). Thus, short-term IGF-I infusion in the lamb can both reduce protein degradation and increase protein synthesis in a number of different tissues. In long-term studies, well-fed yearling sheep were treated for 8 weeks with IGF-I, which increased plasma concentrations of IGF-I by approximately 40–60% (Cottam et al., 1992). There was no increase in body weight or food intake. The weight of the spleen increased by 40%, but there was no effect of IGF-I on carcass weight, composition and dimensions, or on long bone length, although the weight per unit length of the tibia was increased. The somatogenic effect of long-term IGF-I treatment was minimal, suggesting that in well-fed animals with an intact somatotropic axis, IGF-I treatment at doses which increase plasma IGF-I within the physiological range do not enhance somatic growth performance (Cottam et al., 1992). Exogenous IGF-I may suppress tissue IGF-I production by interactive feedback between the systemic and the local IGF-I system. Of particular note in the long-term IGF-I treatment study were the effects on plasma glucose and insulin (Cottam et al., 1992). There was a marked and rapid decline in circulating levels of insulin. Decreased plasma insulin was associated with, and may have been responsible for, a significant increase in blood glucose levels in IGF-I-treated sheep. It is possible that insulinopenia reduces the growth response by restricting substrate uptake. Thus, the insulinopenic effect of long-term IGF-I treatment contrasts with the effect of GH, which induces hyperinsulinaemia. The lack of a somatogenic effect with long-term IGF-I treatment may also be related to changes in IGFBPs. While bGH treatment of sheep significantly increases plasma IGFBP-3 and the ternary 150 kDa complex, IGF-I treatment of sheep was unable to induce such an increase. Since plasma IGFBP-1 was increased and IGFBP-2 levels were suppressed with IGF-I treatment, the change in the relative concentration of the different IGFBPs may have affected the availability of injected IGF-I to target tissues. An independent but related study by Min et al. (1996) showed that long-term administration of IGF-I in young energy-restricted sheep lead to negative feed-back of circulating IGF-I on GH secretion and the GHR number in the liver. Activation of such negative feed-back loops by IGF-I treatment explains the lack of effects on growth rate, metabolic parameters or body composition in endocrinologically normal animals.

Role of the somatotropic axis in conditions of metabolic imbalance The somatotropic axis plays an important role in maintaining homeostasis under a variety of stress conditions including malnutrition, starvation, microbial infection and septic shock. During critical illness body protein and energy reserves are used at a greatly accelerated rate which leads to severe muscle wasting, impaired wound healing, reduced cardiopulmonary performance and diminished immune function. Both GH and IGF-I treatment have a general protein sparing effect during mild or short-term nutritional restriction in humans and sheep, decreasing protein catabolism and increasing cardiac and skeletal muscle protein synthesis. In humans, GH and IGF-I co-treatment appears to be more beneficial since increased IGFBP-3 ternary complex formation induced by GH reduces the levels of ‘free IGF-I’ and subsequent hypoglycaemia associated with IGF-I treatment alone (Kupfer et al., 1993). However, the ability

Regulation of Growth and Metabolism During Postnatal Development

199

of GH to ameliorate the effects of severe catabolism is much reduced due to the induction of GH resistance in these patients (Roth et al., 1995). It has been argued that in cases such as elective surgery GH pre-treatment prior to the onset of GH resistance could minimize post-operative catabolic stress. Such an approach has been used successfully in the fasted sheep (Ogawa et al., 1996). However, the interactions between the intermediary metabolism, the immune system and the somatotropic axis during infections and septic shock are even more complex. A number of studies have employed endotoxin-induced septic shock to investigate the endocrine interactions under pathophysiological conditions similar to microbial infections. Endotoxin is a complex lipopolysaccharide molecule situated within the outer membrane of Gram-negative bacteria; it is the main mediator of septic shock (Morrison and Ryan, 1987). Endotoxin administration stimulates the immune system and modulates the secretion of several mediators and hormones. Endotoxin induces a rapid increase in plasma concentrations of tumour necrosis factor, interleukin-1, and interleukin-6. Its action is mediated, at least in part, by an endotoxin receptor, which has been identified in brain, pituitary, monocytes and plasma (Holst et al., 1996). Changes in several hormonal axes have also been observed after endotoxin injection; there is an increase in activity of the hypothalamic–pituitary–adrenal axis (Dadoun et al., 1998) and a decrease in thyroid-stimulating hormone and alterations in the gonadotropic axis (Battaglia et al., 1997). Interactions between the immune and endocrine systems are multi-directional and the hormonal changes induced by endotoxin or cytokines can act on the immune system. The pathophysiological responses to endotoxin injection include increased heart rate, increased respiratory rate, changes in lung permeability, activation of the complement cascade, hypoglycaemia, insulinaemia, reduction in plasma concentrations of FFA and hypocalcaemia (Kinsbergen et al., 1994). Septic shock is also associated with metabolic abnormalities, cardiovascular dysfunction, and multiple organ failure mainly related to ischaemia in part due to alterations in regional microcirculatory blood flow (Kinsbergen et al., 1994). Endotoxin administration induces species-specific effects on GH secretion; it increases GH secretion in sheep but decreases it in rats and cattle (Coleman et al., 1993). In a recent study, Hennies et al. (1998) investigated the effects of GH therapy in well-nourished sheep exposed to endotoxin-induced septic shock. Endotoxin injections alone resulted in a modest fall in plasma glucose and no change in plasma insulin concentrations. However, the metabolic responses to bGH (0.3 mg kg⫺1 day⫺1) were markedly altered during endotoxin-induced metabolic stress, leading to a dissociation of the known effects of GH on intermediary metabolic pathways. The anticatabolic effect of GH treatment was abolished despite continuation of considerably elevated levels of plasma IGF-I. Furthermore, the GH-treated sheep showed a decline in plasma insulin and glucose levels within 48 h after endotoxin challenge. This suggests that, after observing an initial phase of diabetogenic effects of GH, endotoxin-induced septic shock altered glucose homeostasis, despite continuation of the marked lipolytic effect of GH. Studies in rats have shown that GH-treatment amplifies the adverse effects of endotoxin-induced septic shock. Rats primed with a GH infusion developed more severe hypoglycaemia, hyperlipidaemia and renal and hepatic dysfunction than endotoxin-treated control animals (Liao et al., 1996). These effects were not observed when IGF-I was administered to rats given an endotoxin challenge, implying that the enhanced toxicity of endotoxin was a direct effect of GH administration. Whether the

B.H. Breier et al.

200

major change in GH action on energy metabolism after endotoxin-induced septic shock in the study of Hennies et al. (1998) reflects changes in glucose up-take and utilization or whether the reduction in plasma glucose levels reflect decreased gluconeogenesis, remains to be investigated. Whilst few studies in different species have shown that GH treatment during conditions of metabolic imbalance may lead to changes in GH action, a recent study by Takala et al. (1999) reports a marked increase in mortality and morbidity among critically ill patients treated with GH. In this study a total of 532 critically ill patients were given GH (at a mean dose of 0.1 mg kg⫺1 body weight day⫺1) or placebo to investigate whether GH therapy can attenuate the catabolic response to injury, major surgery and sepsis. However, the mortality rate in patients receiving GH treatment was two times that in the patients receiving placebo. While the underlying mechanisms of the increased mortality rate in GH-treated critically ill patients are not understood, most of the deaths were attributed to multiple-organ failure and septic shock or uncontrolled infection.

Acknowledgements The authors acknowledge support from the Health Research Council of New Zealand.

References Adamo, M.L., Lanau, F., Neuenschwander, S., Werner, H., LeRoith, D. and Roberts, C.T. (1993) Distinct promoters in the rat insulin-like growth factor-I (IGF-I) gene are active in CHO cells. Endocrinology 132, 935–937. Arany, E., Zabel, P., Freeman, D. and Hill, D.J. (1993) Elimination of radiolabelled recombinant human insulin-like growth factor binding protein-3 from the circulation, and its distribution amongst organs and tissues in adult male rats. Regulatory Peptides 48, 133–143. Baker, J., Liu, J.-P., Robertson, E.J. and Efstratiadis, A. (1993) Role of insulin-like growth factors in embryonic and postnatal growth. Cell 75, 73–82. Battaglia, D.F., Bowen, J.M., Krasa, H.B., Thrun, La., Viguie, C. and Karsch, F.J. (1997) Endotoxin inhibits the reproductive neuroendocrine axis while stimulating adrenal steroids: a simultaneous view from hypophyseal portal and peripheral blood. Endocrinology 138, 4273–4281. Baumann, G. (1994) Growth hormone-binding proteins: state of the art. Journal of Endocrinology 141, 1–6. Baxter, R.C. (1995) Insulin-like growth factor binding proteins as glucoregulators. Metabolism 44 (Suppl. 4), 12–17. Bingham, B., Oldham, E.R. and Baumbach, W.R. (1994) Regulation of growth hormone receptor and binding protein expression in domestic species. Proceedings of the Society for Experimental Biology and Medicine 206, 195–199. Blair, H.T., McCutcheon, S.N., Mackenzie, D.D.S., Ormsby, E., Siddiqui, R.A., Breier, B.H. and Gluckman, P.D. (1988) Genetic selection for insulin-like growth factor-I in growing mice is associated with altered growth. Endocrinology 123, 1690–1692. Boge, A., Sauerwein, H. and Meyer, H.H.D. (1995) IGF-I and insulin receptors in bovine skeletal muscle: Comparisons of different developmental ages, two different genotypes and various individual muscles. Experimental and Clinical Endocrinology and Diabetes 103, 99–104.

Regulation of Growth and Metabolism During Postnatal Development

201

Breier, B.H. and Gluckman, P.D. (1991) The regulation of postnatal growth: Nutritional influences on endocrine pathways and function of the somatotropic axis. Livestock Production Science 27, 77–94. Breier, B.H., Gluckman, P.D. and Bass, J.J. (1988a) Influence of nutritional status and oestradiol-17␤ on plasma growth hormone, insulin-like growth factors-I and -II and the response to exogenous growth hormone in young steers. Journal of Endocrinology 118, 243–250. Breier, B.H., Gluckman, P.D. and Bass, J.J. (1988b) The somatotrophic axis in young steers: influence of nutritional status and oestradiol-17B on hepatic high- and low-affinity somatotrophic binding sites. Journal of Endocrinology 116, 169–177. Breier, B.H., Gluckman, P.D. and Bass, J.J. (1988c) The somatotrophic axis in young steers: influence of nutritional status and oestradiol-17␤ on hepatic high- and low-affinity somatotrophic binding sites. Journal of Endocrinology 116, 169–177. Breier, B.H., Gluckman, P.D., McCutcheon, S.N. and Davis, S.R. (1991) Physiological responses to somatotropin in the ruminant. Journal of Dairy Science 74 (Suppl. 2), 20–34. Breier, B.H., Ambler, G.R., Sauerwein, H., Surus, A. and Gluckman, P.D. (1994) The induction of hepatic somatotrophic receptors after birth in sheep is dependent on parturition associated mechanisms. Journal of Endocrinology 141, 101–108. Chin, E., Zhou, J., Dai, J., Baxter, R.C. and Bondy, C.A. (1994) Cellular localization and regulation of gene expression for components of the insulin-like growth factor ternary binding protein complex. Endocrinology 134, 2498–2504. Coleman, E.S., Elsasser, T.H., Kemppainen, R.J., Coleman, D.A. and Sartin, J.L. (1993) Effect of endotoxin on pituitary hormone secretion in sheep. Neuroendocrinology 58, 111–122. Cottam, Y.H., Blair, H.T., Gallaher, B.W., Purchas, R.W., Breier, B.H., McCutcheon, S.N. and Gluckman, P.D. (1992) Body growth and carcass composition in lambs chronically treated with recombinantly derived insulin-like growth factor-1. Endocrinology 130, 2924–2930. Coverly, J.A. and Baxter, R.C. (1995) Regulation of insulin-like growth factor (IGF) binding protein-3 phosphorylation by IGF-I. Progress in Growth Factor Research 6, 223–229. Dadoun, F., Guillaume, V., Sauze, N., Farisse, J., Velut, J.G., Orsoni, J.C., Gaillard, R. and Oliver, C. (1998) Effect of endotoxin on the hypothalamus pituitary axis in sheep. European Journal of Endocrinology 38,193–197. Davis, S.L., Graf, M., Morrison, C.A., Hall, T.R. and Swift, P.J. (1992) Identification and partial purification of serum growth hormone binding protein in domestic animal species. Journal of Animal Science 70, 773–780. Delhanty, P.J. and Han, V.K. (1993) The expression of insulin-like growth factor (IGF)- binding protein-2 and IGF-II genes in the tissues of the developing ovine fetus. Endocrinology 132, 41–52. Devolder, A., Renaville, R., Sneyers, M., Callebaut, I., Massart, S., Goffinet, A., Burny, A. and Portetelle, D. (1993) Presence of growth hormone-binding proteins in cattle plasma and milk. Journal of Endocrinology 138, 91–98. Douglas, R.G., Gluckman, P.D., Ball, K., Breier, B.H. and Shaw, J.H. (1991) The effects of infusion of insulin-like growth factor (IGF) I, IGF-II, and insulin on glucose and protein metabolism in fasted lambs. Journal of Clinical Investigation 88, 614–622. Elsasser, T.H., Rumsey, T.S. and Hammond, A.C. (1989) Influence of diet on basal and growth hormone-stimulated plasma of IGF-I in beef cattle. Journal of Animal Science 67, 128–141. Fowden, A.L., Szemere, J., Hughes, P., Gilmour, R.S. and Forhead, A.J. (1996) The effects of cortisol on the growth rate of the sheep fetus during late gestation. Journal of Endocrinology 151, 97–105. Gallaher, B.W., Breier, B.H., Oliver, M.H., Harding, J.E. and Gluckman, P.D. (1992) Ontogenic differences in the nutritional regulation of circulating IGF binding proteins in sheep plasma. Acta Endocrinologica 126, 49–54. Gallaher, B.W., Oliver, M.H., Eichhorn, K., Kessler, U., Kiess, W., Harding, J.E., Gluckman,

202

B.H. Breier et al. P.D. and Breier, B.H. (1994) Circulating insulin-like growth factor binding proteins (IGFBPs) and IGF-II/mannose-6-phosphate receptor levels are regulated by glucose and insulin in the fetal sheep. Acta Endocrinologica 131, 398–404. Gallaher, B.W., Breier, B.H., Blum, W.F., McCutcheon, S.N. and Gluckman, P.D. (1995) An homologous radioimmunoassay for ovine insulin-like growth factor binding protein 2 (oIGFBP-2): ontogenesis and the response to growth hormone, placental lactogen and IGFI treatment. Journal of Endocrinology 144, 75–82. Gallaher, B.W., Breier, B.H., Keven, C.L., Harding, J.E. and Gluckman, P.D. (1998) Fetal programming of insulin-like growth factor (IGF)-1 and IGF-binding protein-3: evidence for an altered response to undernutrition in late gestation following exposure to periconceptual undernutrition in the sheep. Journal of Endocrinology 159, 501–508. Gatford, K.L., Egan, A.R., Clark, I.J. and Owens, P.C. (1998) Sexual dimorphism of the somatotrophic axis. Journal of Endocrinology 157, 373–389. Gilmour, R.S. (1994) The implications of insulin-like growth factor mRNA heterogeneity. Journal of Endocrinology 140, 1–3. Giudice, L.C., van Dessel, H.J.H.M., Cataldo, N.A., Chandrasekher, Y.A., Yap, O.W.S. and Fauser, B.C.J.M. (1995) Circulating and ovarian IGF binding proteins: potential roles in normo-ovulatory cycles and in polycystic ovarian syndrome. Progress in Growth Factor Research 6, 397–408. Hammerman, M.R. and Miller, S.B. (1993) The growth hormone insulin-like growth factor axis in kidney revisited. American Journal of Physiology 265, F1–F14. Hennies, M., Breier, B.H., Gallaher, B.W., Krishnamurthi, R.V. and Gluckman, P.D. (1998) Growth hormone induces electrolyte and metabolite imbalance in endotoxin challanged sheep. Proceedings of the 80th Annual Meeting of the Endocrine Society, New Orleans, USA, pp. 1–108. Holst, O., Ulmer, A.J., Brade, H., Flad, H.D. and Rietschel, E.T. (1996) Biochemistry and cell biology of bacterial endotoxins. FEMS Immunology Medicine Microbiology 16, 83–104 Holt, R.I., Baker, A.J., Jones, J.S., Crossey, P.A., Stone, N.M., Preedy, V.R. and Miell, J.P. (1996) Differential effects of malnutrition, bile duct ligation and galactosamine injection in young rats on serum levels and gene expression of IGF-binding proteins. Journal of Endocrinology 149, 465–472. Hooper, S.B., Bocking, A.D., White, S.E., Fraher, L.J., McDonald, T.J. and Han, V.K.M. (1994) Catecholamines stimulate the synthesis and release of insulin-like growth factor binding protein-1 (IGFBP-1) by fetal sheep liver in vivo. Endocrinology 134, 1104–1112. Hua, K.M., Hodgkinson, S.C. and Bass, J.J. (1995) Differential regulation of plasma levels of insulin-like growth factors-I and -II by nutrition, age and growth hormone treatment in sheep. Journal of Endocrinology 147, 507–516. Ilondo, M.M., Damholt, A.B., Cunningham, B.A., Wells, J.A., Demeyts, P. and Shymko, R.M. (1994) Receptor dimerization determines the effects of growth hormone in primary rat adipocytes and cultured human IM-9 lymphocytes. Endocrinology 134, 2397–2403. Isaksson, O.G.P., Lindahl, A., Isgaard, J., Nilsson, A., Tornell, J. and Carlsson, B. (1991) Dual regulation of cartilage growth. In: Spencer, E.M. (ed.) Modern Concepts of Insulin-Like Growth Factors. Elsevier, Amsterdam, pp. 121–126. Jones, J.I. and Clemmons, D.R. (1995) Insulin-like growth factors and their binding proteins: biological actions. Endocrine Reviews 16, 3–34. Kiess, W., Yang, Y., Kessler, U. and Hoeflich, A. (1994) Insulin-like growth factor II (IGF-II) and the IGF-II/ mannose-6-phosphate receptor: The myth continues. Hormone Research 41, 66–73. Kinsbergen, M., Bruckmaier, R.M. and Blum, J.W. (1994) Metabolic, endocrine and haematological responses to intravenous E. coli endotoxin administration in 1-week-old calves. Zentralblatt fur Veterinärmedizin – Reihe A 41, 530–547.

Regulation of Growth and Metabolism During Postnatal Development

203

Klempt, M., Bingham, B., Breier, B.H., Baumbach, W.R. and Gluckman, P.D. (1993) Tissue distribution and ontogeny of growth hormone receptor mRNA and ligand binding to hepatic tissue in the midgestation sheep fetus. Endocrinology 132, 1071–1077. Koea, J.B., Breier, B.H., Shaw, J.H. and Gluckman, P.D. (1992a) A possible role for IGF-II: Evidence in sheep for in vivo regulation of IGF-I mediated protein anabolism. Endocrinology 130, 2423–2425. Koea, J.B., Gallaher, B.W., Breier, B.H., Douglas, R.G., Hodgkinson, S., Shaw, J.H.F. and Gluckman, P.D. (1992b) Passive immunization against circulating IGF-I increases protein catabolism in lambs: Evidence for a physiological role for circulating IGF-I. Journal of Endocrinology 135, 279–284. Kupfer, S.R., Underwood, L.E., Baxter, R.C. and Clemmons, D.R. (1993) Enhancement of the anabolic effects of growth hormone and insulin-like growth factor-i by use of both agents simultaneously. Journal of Clinical Investigation 91, 391–396. Lee, P.D., Conover, C.A. and Powell, D.R. (1993) Regulation and function of insulin-like growth factor-binding protein-1. Proceedings of the Society for Experimental Biology and Medicine 204, 4–29. Lewitt, M.S., Scott, F.P., Clarke, N.M. and Baxter, R.C. (1995) Developmental regulation of circulating insulin-like growth factor binding proteins in normal pregnancies and in preeclampsia. Progress in Growth Factor Research 6, 475–480. Li, J., Saunders, J.C., Gilmour, R.S., Silver, M. and Fowden, A.L. (1993) Insulin-like growth factor-II messenger ribonucleic acid expression in fetal tissues of the sheep during late gestation: effects of cortisol. Endocrinology 132, 2083–2089. Li, J., Owens, J.A., Owens, P.C., Saunders, J.C., Fowden, A.L. and Gilmour, R.S. (1996) The ontogeny of hepatic growth hormone receptor and insulin-like growth factor-I gene expression in the sheep fetus during late gestation: developmental regulation by cortisol. Endocrinology 137, 1650–1657. Liao, W., Rudling, M. and Angelin, B. (1996) Growth hormone potentiates the in vivo biological activities of endotoxin in the rat. European Journal of Clinical Investigation 26, 254–258. Liu, F., Powell, D.R., Styne, D.M. and Hintz, R.L. (1991) Insulin-like growth factors (IGFs) and IGF-binding proteins in the developing rhesus monkey. Journal of Clinical Endocrinology and Metabolism 72, 905–911. McCusker, R.H., Cohick, W.S., Busby, W.H. Jr. and Clemmons, D.R. (1991) Evaluation of the developmental and nutritional changes in porcine insulin-like growth factor-binding protein-1 and -2 serum levels by immunoassay. Endocrinology 129, 2631–2638. Min, S.H., Mackenzie, D.D.S., Breier, B.H., McCutcheon, S.N. and Gluckman, P.D. (1996) Responses of young energy-restricted sheep to chronically administered IGF-I: Evidence that IGF-I suppresses the hepatic growth hormone and receptor. Endocrinology 137, 1129–1137. Min, S.H., Mackenzie, D.D.S., Breier, B.H., McCutcheon, S.N. and Gluckman, P.D. (1999) Ontogeny of insulin-like growth factor-1 responsiveness to bovine growth hormone in young lambs. Domestic Animal Endocrinology 17, 409–419. Morrison, D.C. and Ryan, J.L. (1987) Endotoxins and disease mechanisms. Annual Reviews in Medicine 38, 417–432. Ogawa, E., Breier, B.H., Bauer, M.K., Gallaher, B.W., Grant, P.A., Walton, P.E., Owens, J.A. and Gluckman, P.D. (1996) Pretreatment with bovine growth hormone is as effective as treatment during metabolic stress to reduce catabolism in fasted lambs. Endocrinology 137, 1242–1248. Olchovsky, D., Song, J., Gelato, M.C., Sherwood, J., Spatola, E., Bruno, J.F. and Berelowitz, M. (1993) Pituitary and hypothalamic insulin-like growth factor-I (IGF-I) and IGF-I receptor expression in food-deprived rats. Molecular and Cellular Endocrinology 93, 193–198. Pell, J.M., Saunders, J.C. and Gilmour, R.S. (1993) Differential regulation of transcription initiation from insulin-like growth factor-I (IGF-I) leader exons and of tissue IGF-I expression

204

B.H. Breier et al. in response to changed growth hormone and nutritional status in sheep. Endocrinology 132, 1797–1807. Roth, E., Valentini, L., Semsroth, M., Holzenbein, T., Winkler, S., Blum, W.F., Ranke, M.B., Schemper, M., Hammerle, A. and Karner, J. (1995) Resistance of nitrogen metabolism to growth hormone treatment in the early phase after injury of patients with multiple injuries. Journal of Trauma 38, 136–141. Rotwein, P. (1991) Structure, evolution, expression and regulation of insulin-like growth factors I and II. Growth Factors 5, 3–18. Roupas, P. and Herington, A.C. (1994) Postreceptor signalling mechanisms for growth hormone. TEM 5, 154–158. Sauerwein, H., Breier, B.H., Bass, J.J. and Gluckman, P.D. (1991) Chronic treatment with bovine growth hormone upregulates high-affinity hepatic somatotropic receptors in sheep. Acta Endocrinologica 124, 1–7. Scharf, J.G., Ramadori, G., Braulke, T. and Hartmann, H. (1995) Cellular localisation and hormonal regulation of insulin-like growth factor binding proteins and of the acid-labile subunit within rat liver. Progress in Growth Factor Research 6, 175–180. Sherlock, R.G., Blair, H.T., Kuhn, B., Mackenzie, D.D.S., Breier, B.H., Gluckman, P.D. and McCutcheon, S.N. (1993) Selection for insulin-like growth factor-1 causes hypoglycaemia in sheep. Proceedings of the New Zealand Society of Animal Production 53, 381–384. Smith, W.J., Underwood, L.E. and Clemmons, D.R. (1995) Effects of caloric or protein restriction on insulin-like growth factor-I (IGF-I) and IGF-binding proteins in children and adults. Journal of Clinical Endocrinology and Metabolism 80, 443–449. Soos, M.A., Field, C.E., Nave, B.T. and Siddle, K. (1994) Hybrid and atypical insulin-like growth factor receptors. In: Baxter, R.C., Gluckman, P.D. and Rosenfeld, R.G. (eds) The Insulin-like Growth Factors and their Regulatory Proteins. Elsevier, Amsterdam, pp. 95–106. Takala, J., Ruokonen, E., Webster, N.R., Nielsen, M.S., Zandstra, D.F., Vundelinckx, G. and Hinds, C.J. (1999) Increased mortality associated with growth hormone treatment in critically ill adults. The New England Journal of Medicine 341, 785–792. Thomas, G.B., Cummins, J.T., Sudbury, A.W., McCloud, P.I. and Clarke, I.J. (1991) Effect of restricted feeding on the relationship between hypophysial portal concentrations of growth hormone (GH)-releasing factor and somatostatin, and jugular concentrations of GH in ovariectomized ewes. Endocrinology 128 1151–1158.

12

Direct Effects of Photoperiod on Lipid Metabolism, Leptin Synthesis and Milk Secretion in Adult Sheep

Y. CHILLIARD AND F. BOCQUIER Adipose Tissue and Milk Lipids Team, Herbivore Research Unit, INRA-Theix, 63122 St Genès Champanelle, France

Introduction Grazing ruminants have to cope with large seasonal changes in the amount and quality of herbage consumed. Their ability to survive is improved by seasonality of reproduction, which confines productive activities, namely late gestation, lactation and growth, to spring and summer when the climate is favourable and food of good quality is abundant. Furthermore, the chances of survival are increased by depositing fat reserves which are then used during periods of food shortage. Domesticated livestock, which are descended from ancestral types that were probably strongly seasonal, may have conserved some of these peculiarities. However, genetic selection pressure, applied under standardized conditions and aimed at increasing year-round production, may have had the effect of smoothing this seasonality. Sheep, however, are not able to reproduce during all seasons. Nevertheless, there is great variability between sheep breeds, the most sensitive to seasonal changes being located in high latitudes (Soay breed of the Hebrides), while tropical breeds exhibit few responses to seasonal changes. Among seasonal cues, photoperiod (i.e. daily light duration) has been clearly identified to be an effective signal which controls the period of reproduction in sheep (Ortavant et al., 1988; Lincoln and Richardson, 1998). This raises the possibility that photoperiod, which varies during the year, may also have effects during subsequent physiological stages, i.e. reinforcing other physiological adaptive mechanisms.

Food intake and energy expenditure Voluntary food intake is generally linked to physiological status and the corresponding level of animal performance (growth rate, milk yield). Because of interactions between photoperiod and these factors, only results obtained with non-pregnant, non-lactating adult sheep will be discussed in this section. Voluntary food intake is increased (45%) by long daylength exposure in wethers (Gordon, 1964). This effect is more marked © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

205

206

Y. Chilliard and F. Bocquier

with high-quality diets (+11% with fescue hay and +35 % with dehydrated grass; Milne et al., 1978). There are, however, differences between rams of Soay (+71%) vs. Suffolk (+52%) breeds, and gonad-intact vs. castrated animals, the former being more sensitive (Kay, 1985). Although it is generally accepted that minimum food intake occurs during winter and maximal food intake during summer, few experiments have been conducted to analyse the exact timing of the effect of photoperiod on food intake. When photoperiod variations are cyclic (semestrial cycle), maximum food intake occurs between 6 and 13 weeks after the maximum duration of light is reached (Kay, 1985). However, in the case of abrupt changes of photoperiod, the food intake of Soay rams begins to change within 2 weeks (Lincoln and Richardson, 1998). In sheep energy balance studies, neither the digestibility (Walker et al., 1991) nor the metabolizability (Blaxter and Boyne, 1982) of the diet were affected by daylength. Voluntary food intake is generally positively correlated with fasting heat production, and this could suggest that intake oscillations may be driven by seasonal changes in metabolic rate. To test this hypothesis, measurements of heat production were made at a constant level of food intake (near maintenance) in intact ewes subjected to the following photoperiod treatments: constant long-day, natural, simulated natural or reverse-natural (Walker et al., 1991). Oscillations of metabolic heat production observed in ewes subjected to constant illumination provide evidence of a long-term (more than one year) endogenous rhythm in energy metabolism. In all other sinusoidal photoperiodic treatments, whole animal energy metabolism was driven by the pattern of changes in daylight, even when the light cycle was reversed. The peak of metabolic heat production did not occur in phase with the longest or the shortest daylength periods, but 15 ± 3 weeks following the start of the day-length increment (i.e. 11 weeks before the maximum duration of light). The amplitude of changes in fasting heat production (+21%) is in general agreement with previous results obtained in mature wethers (+29%) kept under natural daylength (Blaxter and Boyne, 1982). It is noteworthy that the relative annual amplitude of variation in metabolic rate is far lower than that of food intake variations. Furthermore, maximum heat production under annual photoperiod conditions occurs before maximum daylength and long before maximum food intake would occur.

Adipose tissue and muscle metabolism Wild ruminants exhibit large adaptations to the seasonality of environmental changes. Detailed studies of seasonal changes in adipose tissue metabolism have been done in the reindeer, which inhabits arctic areas. Interestingly, seasonal body weight changes of this animal are mainly due to body fat, and the non-fat component remains fairly constant (Reimers et al., 1982). This suggests a high capacity to spare body proteins, associated with extensive use of body fat during the winter period. The reindeer shows a marked reduction in voluntary food intake during winter, even when offered the same diet ad libitum throughout the year (Larsen et al., 1985b). This results in seasonal cycles of fat deposition and adipose tissue lipogenic activity which parallel those of food intake and are accompanied by changes in plasma concentrations of glucose, acetate, triglyceride, cholesterol and insulin. There was no seasonal change in the concentration of plasma catecholamines, although plasma non-esterified fatty acids were higher dur-

Direct Effects of Photoperiod on Lipid Metabolism

207

ing the winter period (Larsen et al., 1985a,b). However, a decrease in in vitro adrenaline-stimulated lipolytic capacity was observed during winter. This could prevent excessive breakdown of fat stores. In the domestic sheep in Scotland, the rate of adipose tissue fatty acid synthesis and the activity of the lipoprotein lipase, a key enzyme for tissue uptake of plasma triglycerides, increased between October and May (Vernon et al., 1986). These increases were, however, probably related to increased food intake, since plasma concentrations of insulin, glucose and acetate were increased also. In the animal models discussed previously, changes in adipose tissue metabolism paralleled changes in voluntary food intake and/or changes in plasma insulin and energy metabolites that were available for lipogenesis. For this reason it is not possible to distinguish whether the effects were due to a direct effect of photoperiod (or other seasonal factors) on adipose tissue metabolism or due to indirect effects related to changes in food intake. In order to avoid these problems Bocquier et al. (1998) used dry non-pregnant adult ewes which were subjected during winter to either long (16L:8D) or short (8L:16D) daylength for 4–6 weeks, and given fixed amounts of food within two feeding treatments: (i) restricted to 22% of energy requirements for 1 week; or (ii) subsequently re-fed to 190% of requirements for 2 weeks (pair-feeding across photoperiod treatments). Ewes were also ovariectomized in order to avoid any putative indirect effect of photoperiod via changes in the secretion of reproductive hormones, as these are potential modulators of tissue metabolism (Chilliard, 1987). Under these conditions, long days increased lipoprotein lipase activity in Longissimus thoracis muscle and, in overfed ewes, the activities of lipoprotein lipase and malic enzyme (an enzyme involved in NADPH generation for fatty acid synthesis) in subcutaneous adipose tissue, as well as malic enzyme activity in perirenal adipose tissue. Furthermore, long days increased the amount of lipoprotein lipase mRNA in cardiac muscle and perirenal adipose tissue (Table 12.1). There was also a non-significant trend for long days to slightly increase the activity of three other lipogenic enzymes in adipose tissues of re-fed ewes (Bocquier et al., 1998; Faulconnier et al., 1999). Simultaneously, long days increased plasma leptin concentrations and the amount of adipose tissue leptin mRNA independent of feeding level, and decreased plasma non-esterified fatty acids concentrations in underfed ewes (Fig. 12.1). It was previously reported that long photoperiod also increased adipose tissue leptin mRNA in Djungarian hamsters (Klingenspor et al., 1996). However, food consumption was not controlled in this trial. Furthermore, adipose tissue mass and gonadal activity, which have been shown to modulate leptin production in rats and humans, were sharply increased along with leptin mRNA (Klingenspor et al., 1996). The study of Bocquier et al. (1998) in sheep shows, for the first time, an effect of photoperiod on leptin that is independent of food intake, adiposity or gonadal activity. Moreover, the photoperiod-driven changes in leptin, lipogenic enzyme activities and plasma non-esterified fatty acids were not related to changes in plasma insulin, glucose, acetate, lactate, 3-hydroxybutyrate, triglycerides or urea concentrations, since these parameters were not affected by photoperiod. These results suggest the existence of direct effects of photoperiod on sheep adipose tissue lipogenic potential, leptin secretion and lipomobilization, as well as on muscle lipoprotein lipase activity, which are not induced by nutrient supply.

208

Y. Chilliard and F. Bocquier

Table 12.1. Effects of nutritional status and daylength on malic enzyme activity in adipose tissue and lipoprotein lipase activity and mRNA in adipose tissue and muscles (from Bocquier et al., 1998; Faulconnier et al., 1999). Underfed

Re-fed

Short

Long

Short

Long

Statistical effectsa

Subcutaneous adipose tissue Malic enzyme activityb Lipoprotein lipase activityc

41 34

51 35

117 264

158 415

F, p F, P, F⫻P

Perirenal adipose tissue Malic enzyme activity Lipoprotein lipase activityc Lipoprotein lipase mRNAd

74 158 22

87 184 32

147 589 111

208 604 291

F, P, f⫻p F F, p

Cardiac muscle Lipoprotein lipase activitye Lipoprotein lipase mRNAd

133 71

129 118

187 158

160 252

F F, p

Longissimus thoracis muscle Lipoprotein lipase activitye

23

30

33

43

F,p

a5

ewes in each group (N = 20). P, p: effect of photoperiod (P: P < 0.05; p: P < 0.10). F, f: effect of feeding level (F: P < 0.05; f: P < 0.10). f⫻p: = interaction (P < 0.10). b nmol NADPH min⫺1 10⫺6 adipocytes. c nmol fatty acids min⫺1 10⫺6 adipocytes. d Arbitrary units. e nmol fatty acids min⫺1 µg⫺1 DNA.

Secretion of milk and nutrient partitioning Linzell (1973) demonstrated that the rate of milk secretion oscillates in a sinusoidal fashion when goats are kept in continuous milk production for 2–4 years, with maximum milk yields occurring in summer and minimum yields in winter. This phenomenon was confirmed in non-cyclic ovariectomized goats and was not diet related. In selected dairy cows there is a limited effect of photoperiod on milk yield (+5 to +10%), which is followed by an increase in food intake (e.g. Dahl et al., 1997). In sheep there is a clear influence of long photoperiod on milk secretion during early lactation (+52% and +30% in two trials with Prealpes ewes, Bocquier et al., 1986; +25% in Sardinian ewes, Bocquier et al., 1997). Furthermore, the effect of long photoperiod was proportional to the time elapsed between full establishment of daylength and lambing. Large responses (52 and 30%) were obtained when photoperiod treatments started 42 days before lambing, whereas a smaller response (25%) was observed when treatments were established 25 days before lambing. In addition, no effect was detected when photoperiod was changed within the week prior to lambing (Bocquier et al., 1990). A progressive reversal of long and short photoperiod treatments during established lactation altered lactation persistency and led to large differences in favour of long photoperiod (+12 to 34%; Fig. 12.2). In order to separate the effects of constant

Direct Effects of Photoperiod on Lipid Metabolism

209

Fig. 12.1. Effects of daylength (shaded bars: short days (8 h day⫺1) and open bars: long days (16 h day⫺1)) and nutritional status (underfed (22%) or overfed (190% of energy requirements)) on (a) plasma leptin, (b) adipose tissue leptin mRNA (in arbitrary units), (c) plasma NEFA and (d) plasma prolactin (5 ewes per group) (from Bocquier et al., 1998).

light duration from those of changes in light duration, light treatments were abruptly switched from constant long (15.30L:08.30D), or constant short (08.30L:15.30D) to a common treatment of equal light and dark phases (12.00L:12.00D) (Bocquier et al., 1997). The shortening (−3.30 h) of light exposure induced a dramatic decline in milk yield (−38% 42 days later), while lengthened light exposure (+3.30 h) limited the normal decline of milk yield to 8% during the same period. Hence sheep milk production is affected both by changes in daylength and by the duration of light exposure itself, via multiple effects either on mammogenesis, lactogenesis and/or galactopoiesis. Milk fat and protein concentrations are lower during long than during short days. During early lactation, the maximal difference between short and long days was observed (Bocquier, 1985; Bocquier et al., 1997) 30–40 days after lambing for both fat (Prealpes: −10 g l⫺1; Sardinian: −14 g l⫺1) and protein content (Prealpes: −3 g l⫺1; Sardinian: −11 g l⫺1). After the progressive switch of light treatments during late lactation (Fig. 12.2), milk concentration of ewes exposed to long days was even more lowered (−19 g l⫺1 for fat and −13 g l⫺1 for protein) when compared with ewes exposed to short days. After the abrupt change in daylength to an equal dark/light exposure treatment (Bocquier et al., 1997), it took 23 days until the maximal difference in milk concentration occurred. During this period, fat and protein content of milk from ewes subjected to an increase in daylength (+3.30 h) decreased by 8 g l⫺1 and 6 g l⫺1 respectively compared with that from ewes subjected to a decrease (−3.30 h) in daylength.

210

Y. Chilliard and F. Bocquier

Fig. 12.2. Milk production of ewes subjected to either long (嘷) or short (●) photoperiod; in trial A (January lambing; n = 24 per group) and trial B (September lambing; n = 22 per group). Light treatments were established 42 days before lambing, and were reversed from the 9th week of lactation, by progressive changes within a fortnight ( ) (Bocquier, 1985).

This suggests that a single change from long or short days towards a 12L:12D constant photoperiod is sufficient to produce responses in milk yield and composition that could be as high as 50% of the response attained after an exposure of several weeks to a constant difference of 8 h in daily light exposure. The negative relationship between the level of milk yield, and milk fat and protein concentrations may be due to a dilution effect since the increase in milk yield by long daylength (+20% in early lactation, i.e. between day 1 and day 30, and +67% in late lactation, i.e. between day 90 and day 110; Table 12.2) is accompanied by a decline in both fat and protein content of milk (−5% in early lactation and −19% in late lactation). This decline is, however, lower than the positive effect on milk yield, because both fat and protein yields are increased by long daylength in early (+14%) and late (+36%) lactation (Table 12.2). Surprisingly, the effects of photoperiod on milk lactose and mineral content have not been studied, despite the well-known contribution of these constituents to the osmotic regulation of water secretion in milk. In ad libitum-fed ewes, voluntary intake was identical between groups during the first month of lactation and became higher (+16% after 150 days of lactation) in longday exposed ewes (Bocquier et al., 1997). This resulted in a lower calculated energy balance (−11% of energy requirements) during the first month of lactation for long-day

Direct Effects of Photoperiod on Lipid Metabolism

211

Table 12.2. Photoperiod and milk secretion by dairy ewes (Bocquier, 1985). Early lactationa

Late lactationb

Shortc

Longd

Long % Short

Shortc

Longd

Long % Short

980

1178

120

328

550

167

Content (g kg ) Fat Protein

73 50

69 47

95 94

99 71

80 57

81 80

Yield (g day⫺1) Fat Protein

71 49

82 56

114 114

32 22

43 31

135 137

⫺1

Milk yield (g day ) ⫺1

a From

1 to 30 days after lambing, data from trials A and B in Fig. 12.2. 91 to 110 days after lambing. c 8 h day⫺1. d 16 h day⫺1. b From

exposed ewes which produced more milk energy, than for short-day exposed ewes (−5% of requirements), while the reverse was observed between days 30 and 150 of lactation (+22% and +16% of requirements, for long- and short-day exposed ewes, respectively). In order to isolate a direct effect of photoperiod on milk secretion Bocquier et al. (1986) fed ewes a limited and identical amount of food in both photoperiod treatments. Under these conditions, there was an increase in the apparent efficiency of milk yield at the expense of body energy deposition in ewes exposed to long days (Fig. 12.3). This partitioning of energy towards milk output occurred in ewes that were either in positive or equilibrated energy balance. Changes in body energy content were mainly due to changes in body lipids, suggesting that, in sheep, photoperiod acts on the partitioning of nutrients between mammary gland and adipose tissue independently of effects on food intake.

Hormonal mediation of photoperiodic changes Brain Secretion of plasma melatonin from the pineal gland, which is lowest during the light and greatest during the dark period, is considered as the hormonal signal related to the length of day. In sheep, pinealectomy greatly reduced the photoperiod-induced changes in food intake, growth rate and reproduction (Williams et al., 1995; Malpaux et al., 1996), and melatonin implants, which mimic short days, induced a decline in circulating prolactin (PRL) (Viguié et al., 1997). Similarly, in hamsters, melatonin injections mimicked the effects of short days on body lipids, and these effects were initiated before any change in food intake. Pinealectomy, however, blocked the effects of short days on reproduction but not on body weight, suggesting that retinal photoreception could be involved in body weight regulation (Bartness and Wade, 1984). Considering the fact that sheep maintained in constant photoperiod eventually become cyclic,

212

Y. Chilliard and F. Bocquier

Fig. 12.3. Effect of photoperiod (long versus short) on energy partitioning between milk (open bars) and body reserves (shaded bars) in lactating (pair-fed) ewes (Bocquier et al., 1986). Trials A and B are described in Fig. 12.2. Body energy was measured by D20 dilution technique (Bocquier et al., 1999).

suggests that, in the absence of variation in melatonin production, both reproduction (Williams et al., 1995) and metabolic heat production (Walker et al., 1991) are driven by endogenous rhythms. Thus the natural photoperiod may serve to maintain the synchrony between the circannual cycle of physiological activities and the appropriate season (Williams et al., 1995), although melatonin may not be the sole internal signal. The complexity of the transfer of photoperiodic information is shown by the fact that hypothalamo-pituitary disconnection in rams suppresses the cyclical effects of photoperiod changes on α-melanocyte-stimulating hormone (α-MSH), β-endorphin, adrenocorticotrophic hormone (ACTH), luteinizing hormone (LH), follicle-stimulating hormone (FSH), feed intake and body weight, but does not suppress cyclical changes in PRL (Lincoln and Richardson, 1998). This means that photoperiod-driven signals act on pituitary hormones either via the hypothalamus by neurohormonal pathways (α-MSH, β-endorphin, ACTH, LH, FSH) or directly on pituitary secretion (PRL), and that PRL is not directly involved in body weight changes. In view of recent data indicating that leptin acutely stimulates in vitro prolactin release by the rat pituitary (Yu et al., 1997), our results (Fig. 12.1) suggest that the photoperiod- and feeding-induced changes in PRL could be facilitated by leptin.

Peripheral hormones Prolactin is second only to melatonin in terms of responsiveness to photoperiod, and responds to an abrupt change of photoperiod within a week (Lincoln et al., 1978). Prolactin is elevated both in intact (Kann, 1997) and ovariectomized (Bocquier et al.,

Direct Effects of Photoperiod on Lipid Metabolism

213

1998) ewes exposed to long daylength. Long photoperiod is also effective in the pregnant ewe (Perier et al., 1986) and induces a 40% higher and earlier (20 h) pre-partum surge of plasma PRL (Bocquier, 1985). In lactating ewes exposed to long photoperiod, mean plasma PRL concentration was increased by 311% (Bocquier et al., 1990). However, photoperiod-induced increments in serum concentration of PRL began to wane when sheep were maintained on 16L:8D for more than 16 weeks (Almeida and Lincoln, 1984). This indicates that PRL secretion becomes refractory to a prolonged constant photoperiodic stimulus. Plasma growth hormone (GH) has been shown to be slightly elevated by long daylength in rams (Lincoln and Richardson, 1998) and in the pregnant ewe (Perier et al., 1986), but this has never been observed in the lactating ewe (Bocquier, 1985; Bocquier et al., 1990), nor in the dry non-pregnant ovariectomized ewe (Chilliard et al., 1998b). It has been suggested recently that the effect of long photoperiod on milk yield in the cow could be mediated by circulating insulin-like growth factor-I (IGF-I). Cows exposed to long photoperiod produced more milk (+6%) without increased feed intake, and had an elevated concentration of IGF-I (+77%), without altered concentrations of GH (as in lactating sheep) or IGFBP-2 and -3 (Dahl et al., 1997). Other hormones were either unaffected by daylength (insulin; Bocquier et al., 1998) or gave conflicting results (thyroid hormones and corticosteroids; Forbes 1982; Vernon et al., 1986; Petitclerc and Zinn, 1991; Lincoln and Richardson, 1998). In ad libitum-fed sheep, there was an increase in plasma ␤-endorphin concentration at the beginning of a short-day period that could have stimulated insulin secretion and adipose tissue lipogenesis (Lincoln and Richardson, 1998). However, as this followed a period of high food intake at the end of a long-day period, it is not clear whether the changes in ␤-endorphin concentrations were due to photoperiod per se, or to the previous change in food intake.

Adipose tissue Peripheral effects of melatonin on adipose tissue are plausible, since specific binding sites in Siberian hamster brown adipose tissue have been described (Le Gouic et al., 1997). However, the functionality of these receptors remains to be shown. Although plasma PRL concentration increases in response to increasing both daylength and feeding level (Bocquier et al., 1998), this hormone is probably not directly involved in adipose tissue response to daylength because sheep adipose tissue appears to lack PRL receptors (Emane et al., 1986; Knight and Flint, 1995). Furthermore, addition of PRL in vitro did not change lipolysis in bovine or ovine adipose tissue (Houseknecht et al., 1996), nor was lipogenesis changed in ovine (Vernon and Finley, 1988) or bovine adipose tissue (Etherton et al., 1987). There could be, however, indirect effects of PRL in vivo. In the rat, PRL infusion was shown to stimulate the secretion of a mammotrophic and lactogenic hepatic factor (synlactin) (English et al., 1990), the effects of which remain to be unravelled for adipose tissue. PRL injections increased mammary acetyl-CoA carboxylase activity and mRNA abundance in the lactating rat, and decreased it in adipose tissue (Barber et al., 1992). The adipose tissue lipolytic response was increased by PRL injections in female rats during recovery from lactation (Vernon and Finley, 1986). Although these results

214

Y. Chilliard and F. Bocquier

suggest that the effects of PRL in rodents are antilipogenic and/or lipolytic (indirect), there was no in vivo effect of PRL on plasma non-esterified fatty acids in adult sheep (Luthman and Johnson, 1972), lactating dairy cows (Plaut et al., 1987) or lactating goats (Jacquemet and Prigge, 1991). Furthermore, administration of bromocriptine (a dopamine agonist which inhibits PRL secretion) did not change adipose tissue lipogenic rate in post-partum beef cow (Mills et al., 1989), and PRL administration did not affect adipose tissue lipogenic or lipolytic rates, although the weight of perirenal adipose tissue was increased in growing sheep (Eisemann et al., 1984). Results from other studies with bromocriptine suggest that PRL has lipogenic (in synergy with insulin) rather than antilipogenic effects on the liver, and promotes body fat deposition in several rodent species and pig (Cincotta and Meier, 1989a; Cincotta et al., 1989). Interestingly, the decrease in body fat after bromocriptine administration was observed without (or with very few) changes in food consumption, energy expenditure and lean body mass (Cincotta et al., 1993a). However, the antilipogenic effects of bromocriptine could be due to dopaminergic effects in the hypothalamus (Luo et al., 1997), rather than simply due to a decrease in PRL secretion. In rodents, it appears that the lipogenic effects of photoperiod result from a synergy between insulin, corticosteroids and PRL which is dependent on the time of day, both in vivo (Cincotta et al., 1993b) and in vitro (Cincotta and Meier, 1989b). These observations are related to a paradigm which was proposed more than 20 years ago by Meier and Burns (1976) from studies in several vertebrate species (including fishes, birds and mammals) that exhibit marked seasonal cycles in body fat stores: These cycles are controlled by the circadian rhythms of corticosteroids and prolactin. Prolactin stimulates increases or decreases in fat storage depending on whether it is present in larger quantities during daily intervals of lipogenic or lipolytic sensitivities. The intervals of sensitivity are entrained by the daily photoperiod and mediated by the adrenal corticosteroids. Thus the temporal synergism of the circadian rhythms of corticosteroid and prolactin hormones involves a relation between sensitivity rhythms of cells involved in lipid metabolism and rhythms of the stimulatory hormone, prolactin. The cumulative effects of various temporal hormonal patterns account for the seasonal changes in fat stores.

This results in annual cycles of metabolic activities that are the result of an interaction of the daily photoperiod with an endogenous seasonal timing mechanism, termed scotosensitivity. Alteration in PRL rhythm could reflect an altered phase of a central circadian pacemaker that entrains other circadian responses (Meier and Cincotta, 1996). Alternating insulin–epinephrine infusions in the rat resulted in a complete inversion of the normal circadian distribution of sleeping and feeding patterns. This led to the hypothesis that the circadian light cycle might act on sleeping and feeding by influencing the tonus of secretion of metabolic hormones mediating the lipogenic vs. lipolytic ratio (Danguir and Nicolaidis, 1980). However, plasma insulin and GH are not changed by photoperiod in the non-pregnant ewe. Meier’s hypothesis can also be compared with the fact that the in vitro synergism between insulin and corticosteroids for lipogenic effects was more important in sheep than in cattle adipose tissue explants (Chilliard and Faulconnier, 1995; Faulconnier et al., 1996), bearing in mind that the latter species is less photoperiod-sensitive. Nevertheless, the lack of a clear effect of season or photoperiod on plasma insulin and

Direct Effects of Photoperiod on Lipid Metabolism

215

corticosteroid concentrations in sheep does not argue for a simple extrapolation to the ovine species of the PRL–glucocorticoid–insulin hypothesis, although it appears to be applicable to other species. There is a need for information on circadian variations of these three hormones in sheep, the sensitivity of adipose tissue to their effects, and on whether their secretion is altered by variations in daylength. The mechanisms by which photoperiod influences leptin gene expression in adipose tissue remain unknown. It has been shown recently that intravenous injection of neuropeptide Y (NPY) increased leptin mRNA level in ovine adipose tissue (Dyer et al., 1997), but it is not known whether daylength affects NPY or related peptides. It has also been shown that glucocorticoids can modulate leptin in monogastric and ruminant species (Houseknecht et al., 1998; Chilliard et al., 1999). In the Siberian male hamster, exposure to short days activated the sympathetic nervous system and increased norepinephrine turnover (McElroy et al., 1986). A similar catecholaminergic effect of short days, which remains to be demonstrated in sheep, could account for the increase in plasma non-esterified fatty acid concentrations, as well as for decreases in adipose tissue leptin and lipogenic activities (Fig. 12.1 and Table 12.1). Catecholamines are lipolytic and antilipogenic in sheep and it was recently shown that they lowered leptin concentrations in rodents (Trayhurn et al., 1995) and cattle (Chilliard et al., 1998a). Differences in catecholaminergic innervation between anatomical sites of adipose tissues could explain why short days deplete more internal than subcutaneous adipose tissue, a phenomenon which could serve to enhance insulation from the cold during winter (Youngstrom and Bartness, 1995). However, the effect of short days on the pattern of adipose tissue depletion, which was observed in males, was not apparent in female Siberian hamsters (Bartness, 1995). In ovariectomized ewes, short days decreased lipoprotein lipase activity to a greater extent in subcutaneous than in perirenal adipose tissue, and in skeletal than in cardiac muscle (Table 12.1). The noradrenergic sensitivity of adipocytes was not changed in Djungarian hamsters exposed to short days, suggesting that their rapid fat mobilization is determined centrally by the chronic activation of the sympathetic nervous systems, without desensitization of adipocyte ␤-adrenergic lipolysis (Mercer et al., 1995).

Mammary gland The effects of photoperiod on lactation can be analysed either at the level of whole body metabolism and nutrient partitioning or at the level of mammogenesis (i.e. the formation of a functional mammary gland) and galactopoiesis (milk production). The mammogenic effect of long photoperiod observed in ewes can be explained by the elevated and prolonged pre-partum PRL surge (Bocquier, 1985; Perier et al., 1986); prolactin is necessary for the complete structural differentiation of epithelial cells and for the onset of copious milk production (Knight, 1993; Kann, 1997). A role for placental lactogen in photoperiod-induced differences in mammogenesis is, however, doubtful, since circulating concentrations were unchanged by daylength (Perier et al., 1986). Artificial induction of lactation is an interesting model since it avoids the interference with numerous hormones that are prevalent during pregnancy. Prealpes maiden ewes, receiving an oestroprogestative treatment before being submitted to the machine-milking stimulus, produced more milk (+27%) under long-day treatment

216

Y. Chilliard and F. Bocquier

than under short-day treatment (Kann, 1997). In addition to the classical increment of circulating PRL, there was an increase in plasma concentrations of GH (+136%) and IGF-I (+23%), concurrently with an enhancement of GH hepatic receptor number. Although galactopoeitic effects of GH have been clearly demonstrated in cows (Bauman and Vernon, 1993), its mechanism of action remains controversial. GH may not act directly on the mammary gland but via IGF-I. Although the plasma concentration of GH is generally not altered by photoperiod in the lactating sheep, IGF-I was elevated moderately in the ewe (Kann, 1997) and substantially in cattle (Dahl et al., 1997) subjected to long daylength. These observations can be related to an increase in the number of GH receptors in the liver (Kann, 1997). It seems likely then that photoperiod-induced changes in IGF-I concentration may be partly responsible for effects on milk production. Although systematically elevated by long daylength, the exact role of PRL on lactation is not clear, even if this hormone exerts some important galactopoietic effects in the ewe, a smaller effect in the goat (Knight and Flint, 1995) and is without effect in cattle (Plaut et al., 1987). In the ewe it was demonstrated (Hooley et al., 1978; Kann et al., 1978; Gow et al., 1983) that a pharmacological PRL depression, using bromocriptine, decreased milk yield by 20–30%. Furthermore, the negative effects of bromocriptine on milk yield were reversed by concurrent infusion of PRL (Hooley et al., 1978). However, since milk production was not totally suppressed by the almost complete (< 1 ng ml⫺1) depression of PRL, this hormone may not be absolutely necessary in the galactopoeitic hormonal complex of the ewe (Gow et al., 1983). It is worth noting, however, that sheep, which have been less subjected to selection for milk production than dairy cattle, are also more sensitive to the effect of PRL and to the effects of photoperiod, while dairy cattle may have become exquisitely sensitive to small amounts of circulating PRL (Knight, 1993), and hence less dependent on photoperiod because PRL concentration would never be limiting. Separate consideration of the roles of GH or PRL is probably an oversimplification, since it has been shown in goats that GH response is higher when given together with prolactin than when given alone (Knight and Flint, 1995). Although ewes exposed to long daylength had a higher milk-solid yield, their milk was more diluted than that of short-day exposed ewes. One hypothesis is that such a dilution effect may be attributed to high levels of PRL observed in ewes exposed to long daylength. In the case of dairy cattle, injection of exogenous PRL had no effect on milk concentration of fat, protein or lactose and had a small positive effect on α-lactalbumin (Plaut et al., 1987). There is, however, evidence that PRL acts on water and solute transport across mammary epithelia, although no direct effect has been isolated (Sheenan, 1994). The stimulatory effect of photoperiod on milk production is accompagnied by a readjustment of whole body energy metabolism, with a decrease in body fat (Fig. 12.3) before an increase in food intake. The effect of photoperiod may override and exacerbate the homeorhetic (from Greek meaning ‘uniform flow’; Bauman and Currie, 1980) or teleophoretic (from Greek meaning ‘a transport that is oriented towards a goal’; Chilliard, 1986, 1999) mechanisms whereby an animal re-adjusts its homeostatic setpoints to new physiological conditions. Besides the complex interactions linked to the circadian and seasonal rhythms of hormonal secretions and tissue sensitivity to hormones, the effects of PRL on adipose tissue are also dependent on the presence or the

Direct Effects of Photoperiod on Lipid Metabolism

217

absence of a functional lactating mammary gland, as there are antilipogenic or lipolytic effects in lactating rats, but few effects in virgin or pregnant rats (Vernon and Flint, 1983). The mammary gland was, indeed, proposed to act as an endocrine gland (Maule Walker and Peaker, 1978) and thus could play its own role in teleophoretic adaptations which ensure a high priority of nutrient use for lactation.

Conclusion Wild animals have developed numerous non-exclusive adaptive strategies to survive in their environmental conditions. In non-equatorial regions, photoperiod is the most noise-free signal of the annual variations in food resources and general climatic conditions. Logically, the reproductive process is strongly linked to light duration, since it induces the succession of pregnancy and lactation – both of which lead to high nutritional requirements. As the feed resources available to meet this demand are generally maximal in spring and/or summer, the reproduction of the different species occurs at different periods of the year according to gestation length. Because of the wide variation in the length of pregnancy, there must be a great diversity of mechanisms whereby animal species translate the photoperiodic signals into biological adaptations, although there appears to be a strong communality of central neuroendocrine and peripheral signalling systems (melatonin and PRL responses) between species (Morgan and Mercer, 1994). Although domesticated, sheep have not lost the ability to respond to photoperiod. It is of interest to note the differences in time-scale of the maximal response of physiological functions to changes in daylength, namely: milk composition (less than a week), milk production (a few weeks), adipose tissue metabolism (several weeks) and basal metabolic rate and food intake (several months). In parallel with this, hormonal changes are very rapid for prolactin (a few days), rapid for leptin and IGF-I (less than a month), or uncertain (GH, T3, T4, corticosteroids). These results together confirm the hypothesis formulated by Ortavant et al. (1988) that photoperiod may also be involved in metabolic adaptations that accompany the high nutritional requirements for pregnancy and lactation in the sheep. Our proposed interpretation of the effect of photoperiod on metabolic regulation in sheep is illustrated in Fig. 12.4. Homeostasis in the dry non-productive ewe (Fig. 12.4a) reflects regulatory processes that ensure the constancy of internal conditions for vital processes (according to given set-points) in the face of changes in environmental conditions. We speculate that the set-point for body lipids is diminished by short days. Reproduction, which is favoured by short days, may be impaired if ewes are underfed. Hence, the increase in non-esterified fatty acids and decrease in leptin by short days can be seen as facilitating the general adaptation of ewes to winter undernutrition, and increasing the sensitivity of reproduction blockade in these adverse conditions. When the reproductive process is successful, the organism is then subjected to teleophoresis, i.e. the orchestrated control of body tissue metabolism necessary to support physiological functions such as pregnancy or lactation. In the case of the ewe, lactation occurs naturally during long days due to the length of gestation. Here again, the physiological regulation is probably changed because milk yield is increased by daylength at the expense of body reserves (Fig. 12.4b), and later on food intake is also

218

Y. Chilliard and F. Bocquier

(3) Food intake (+) (2) Body fat (–) (1) Milk yield (+)

Feeding level High

Lactation function

(4) Food intake (+) (3) Basal metabolism (+) (2) Leptinaemia (+) (1) Lipogenesis (+)

Teleophoresis Body fat set-point

a Equilibrated

Homeostasis

Homeostasis

Homeostasis Body fat set-point

c

b

(+)

(1) Lipolysis (+) (2) Leptinaemia (–)

(Remnant Reproduction effects)* (+/–)

Insufficient Reproduction

(Yes)

(–)

(–)

(+) (Effects of short days) Dry ewe

(Effects of long days) Pregnant or lactating ewe

Dry ewe

Fig. 12.4. Diagrammatic representation of the effects of photoperiod (䊎) on energy metabolism regulation and body fat set-point during the reproductive cycle in the ewe (* remnant effects due to changes in body fatness during lactation).

increased. After drying off, ewes are generally still naturally subjected to long days (Fig. 12.4c), and body fat is restored towards a set-point that is higher than during the previous short-day season. This is in apparent contradiction with elevated leptin levels, which would decrease food intake and adipose tissue lipogenesis (Houseknecht et al., 1998; Chilliard et al., 1999). As this does not occur, central and peripheral resistance to leptin probably occurs. The sheep subjected to long days could thus furnish a new model to study leptin resistance, a key issue in obesity studies. The biological significance of this ability to restore body reserves during long days is that it occurs at a period when long photoperiod postpones the occurrence of the next reproductive cycle and when an anticipatory body reserve replenishment, when food resource is abundant, is necessary to support future reproduction. Similarly, the increase in muscle lipoprotein lipase activity could reflect an adaptation for sparing body protein despite increasing physical activity and foraging, since the enzyme controls, in part, the entry of energy fuel, as fatty acids, into muscle cells. These observations can be viewed as annual rhythms regulating the expression of the thrifty genotype/phenotype, or fattening physiology, that enabled animals and human ancestors to adapt efficiently to large seasonal changes in food availability and in the nutrient content of food (Meier and Cincotta, 1996).

Direct Effects of Photoperiod on Lipid Metabolism

219

This concept of circannual adaptation through photoperiod re-adjusting set-points for both homeostasis and teleophoresis is in keeping with the views of Mercer (1998) for seasonal mammals, and helps to understand why a ewe placed in phase with optimal daylength is particularly efficient and well adapted to its environment. This setpoint re-adjustment, which results from the photoperiodic rhythms, is anticipatory and preventive rather than corrective as is the homeostatic regulation (Mrosovsky, 1976). The independent effects of photoperiod on reproductive and non-reproductive functions such as cyclicity, milk yield, adipose tissue metabolism, energy expenditure and food intake indicates that sheep metabolism is temporally coordinated by different and complementary (or redundant) mechanisms towards a general adaptation to annual fluctuations in food resources. This is of practical interest when using these animals in extensive farming systems, in which their ability to cope with fluctuations of food resources may be of ecological and economic importance.

Acknowledgements We thank Yannick Faulconnier, Muriel Bonnet, Guy Kann and Daniel Sauvant for helpful discussions during the preparation of the manuscript, and Pascale Béraud for the artwork and secretarial assistance. This research was supported by the INRA ‘Lipogenesis in farm animals’ and ‘Leptin and reproduction in farm animals’ grants.

References Almeida, O.F.X. and Lincoln, G.A. (1984) Reproductive photorefractoriness in rams and accompanied changes in the pattern of melatonin and prolactin secretion. Biology of Reproduction 30, 143–153. Barber, M.C., Travers, T., Finley, E., Flint, D.J. and Vernon, R. (1992) Growth-hormone– prolactin interactions in the regulation of mammary and adipose tissue-tissue acetyl-CoA carboxylase activity and gene expression in lactating rats. Biochemical Journal 285, 469–475. Bartness, T.J. (1995) Short-day-induced depletion of lipid stores is fat pad- and gender-specific in Siberian hamsters. Physiology and Behaviour 58, 539–550. Bartness, T.J. and Wade, G.N. (1984) Photoperiodic control of body weight and energy metabolism in Syrian hamsters (Mesocricetus auratus): Rôle of pineal gland, melatonin, gonads, and diet. Endocrinology 114, 492–498. Bauman, D.E. and Currie, W.B. (1980) Partitioning of nutrients during pregnancy and lactation: a review of mechanism involving homeostasis and homeorhesis. Journal of Dairy Science 63, 1514–1529. Bauman, D.E. and Vernon, R.G. (1993) Effects of exogenous bovine somatotropin on lactation. Annual Review of Nutrition 13, 347–461. Blaxter, K.L. and Boyne, A.W. (1982) Fasting and maintenance metabolism of sheep. Journal of Agricultural Sciences, Cambridge 99, 611–620. Bocquier, F. (1985) Influence de la photopériode et de la température sur certains équilibres hormonaux et sur les performances zootechniques de la brebis en gestation et en lactation. Thèse Doc Ing Sci Agro, Institut National Agronomique Paris-Grignon, 105 pp. Bocquier, F., Thériez, M., Kann, G. and Delouis, C. (1986) Influence de la photopériode sur la partition de l’énergie nette entre la production laitière et les réserves corporelles chez la brebis laitière. Reproduction Nutrition Development 26, 389–390.

220

Y. Chilliard and F. Bocquier Bocquier, F., Kann. G. and Thériez, M. (1990) Relationships between secretory patterns of growth hormone, prolactin and body reserves and milk yield in dairy ewes under different photoperiod and feeding conditions. Animal Production 51, 115–125. Bocquier, F., Ligios, S., Molle, G. and Casu, S. (1997) Effets de la photopériode sur la production, la composition du lait et sur les consommations volontaires chez la brebis laitière. Annales de Zootechnie 46, 427–438. Bocquier, F., Bonnet, M., Faulconnier, Y., Guerre-Millo, M., Martin, P. and Chilliard, Y. (1998) Effects of photoperiod and feeding level on perirenal adipose tissue metabolic activity and leptin synthesis in the ovariectomized ewe. Reproduction Nutrition Development 38, 489–498. Bocquier, F., Guilllouet, P., Barillet, F. and Chilliard, Y. (1999) Comparison of three methods for the in vivo estimation of body composition in dairy ewes. Annales de Zootechnie 48, 297–308. Chilliard, Y. (1986) Revue bibliographique: Variations quantitatives et métabolisme des lipides dans les tissus adipeux et le foie au cours du cycle gestation-lactation. 1. Chez la ratte. Reproduction Nutrition Development 26, 1057–1103. Chilliard, Y. (1987) Revue bibliographique: Variations quantitatives et métabolisme des lipides dans les tissus adipeux et le foie au cours du cycle gestation-lactation. 2. Chez la brebis et la vache. Reproduction Nutrition Development 27, 327–398. Chilliard, Y. (1999) Metabolic adaptations and nutrient partitioning in the lactating animal. In: Martinet, J., Houdebine, L.M. and Head, H.H. (eds) Biology of Lactation. INRA Ed., Paris, France, pp. 503–552. Chilliard, Y. and Faulconnier, Y. (1995) Insulin and(or) dexamethasone regulation of glucose and acetate utilization in ovine and bovine adipose tissue explants incubated for seven days. Journal of Animal Science 73, 2063–2070. Chilliard, Y., Bocquier, F., Delavaud, C., Guerre-Millo, M., Bonnet, M., Martin, P., Faulconnier, Y. and Ferlay, A. (1998a) Leptin in ruminants: effects of species, breed, adiposity, photoperiod, beta-agonists and nutritional status. Proceedings of the Cornell Nutrition Conference for Feed Manufacturers. October 1998, Cornell University, New York, USA, pp. 65–74. Chilliard, Y., Faulconnier, Y., Bonnet, M. and Bocquier, F. (1998b) Mid-term effect of photoperiod on adipose tissue lipogenesis, plasma metabolites, insulin, somatotropin and prolactin in underfed-refed ewes. In: McCracken, K., Unsworth, E.K. and Wylie, A.R.G. (eds) Proceedings of the 14th Symposium on Energy Metabolism of Farm Animals. CAB International, Wallingford, pp. 67–70. Chilliard, Y., Bocquier, F., Delavaud, C., Faulconnier, Y., Bonnet, M., Guerre-Millo, M., Martin, P. and Ferlay, A. (1999) La leptine chez le ruminant. Facteurs de variation physiologiques et nutritionnels. INRA Productions Animales 12, 225–237. Cincotta, A.H. and Meier, A.H. (1989a) Reduction of body fat stores and total plasma cholesterol and triglyceride concentrations in several species by bromocriptine treatment. Life Sciences 45, 2247–2254. Cincotta, A.H. and Meier, A.H. (1989b) Prolactin influences the circadian rhythm of lipogenesis in primary cultured hepatocytes. Hormone and Metabolic Research 21, 64–68. Cincotta, A.H., Meier, A.H. and Southern, L.L. (1989) Bromocriptine alters hormone rhythms and lipid metabolism in swine. Annals of Nutrition and Metabolism 33, 305–314. Cincotta, A.H., MacEachern, T.A. and Meier, A.H. (1993a). Bromocriptine redirects metabolism and prevents seasonal onset of obese hyperinsulinemic state in Syrian hamsters. American Journal of Physiology 264, E285-E293. Cincotta, A.H., Schiller, B.C., Landry, R.J., Herbert, S.J., Miers, W.R. and Meier, A.H. (1993b) Circadian neuroendocrine role in age-related changes in body fat stores and insulin sensitivity of the male Sprague-Drawley rat. Chronobiology International 10, 244–258. Dahl, G.E., Elhasser, T.H., Capuco, A.V., Erdman, R.A. and Peters, R.R. (1997) Effects of a long daily photoperiod on milk yield and circulating concentrations of insulin-like growth factor-I. Journal of Dairy Science 80, 2784–2789.

Direct Effects of Photoperiod on Lipid Metabolism

221

Danguir, J. and Nicolaidis, S. (1980). Circadian sleep and feeding patterns in the rat: possible dependence on lipogenesis and lipolysis. American Physiological Society 238, E223–E230. Dyer, C.J., Simmons, J.M., Matteri, R.L. and Keisler, D.H. (1997) Effects of an intravenous injection of NPY on the leptin and NPY-Y1 receptors mRNA expression in ovine adipose tissue. Domestic Animal Endocrinology 14, 325–333. Eisemann, J.H., Bauman, D.E., Hogue, D.E. and Travis, H.F. (1984) Influence of photoperiod and prolactin on body composition and in vitro lipid metabolism in wether lambs. Journal of Animal Science 59, 95–104. Emane, M.N., Delouis, C., Kelly, P.A. and Djiane, J. (1986) Evolution of prolactin and placental lactogen receptors in ewes during pregnancy and lactation. Endocrinology 118, 695–700. English, D.E., Russel, S.M., Kartz, L.S. and Nicoll, C.S. (1990) Evidence for a role of the liver in the mammotrophic action of prolactin. Endocrinology 126, 2252–2256. Etherton, T.D., Evock, C.M. and Kensinger, R.S. (1987) Native and recombinant bovine growth hormone antagonize insulin action in cultured bovine adipose tissue. Endocrinology 121, 699–703. Faulconnier, Y., Guillon, L. and Chilliard, Y. (1996) Lipoprotein lipase and glucose-6-phosphate dehydrogenase activities in bovine and ovine adipose tissue incubated for 7 days : effects of insulin and/or dexamethasone. Comparative Biochemistry and Physiology 113B, 421–426. Faulconnier, Y., Bonnet, M., Bocquier, F., Leroux, C., Hocquette, J.F., Martin, P. and Chilliard, Y. (1999) Régulation du métabolisme lipidique des tissus adipeux et musculaires chez le ruminant. Effets du niveau alimentaire et de la photopériode. INRA Productions Animales 12, 287–300. Forbes, J.M. (1982) Effects of lighting patterns on growth, lactation and food intake in the sheep, cattle and deer. Livestock Production Science 9, 361–374. Gordon, J.G. (1964) Effect of the time of the year on the roughage intake of housed sheep. Nature 204, 798–799. Gow, C.B., McDowell, G.H. and Jenkin, G. (1983) The importance of prolactin for initiation of lactation in the pregnant ewe. Australian Journal of Biological Science 36, 357–367. Hooley, R.D., Campbell, J.J. and Findlay, J.K. (1978) The importance of prolactin for lactation in the ewe. Journal of Endocrinology 79, 301–310. Houseknecht, K.L., Bauman, D.E., Vernon, R.G., Byatt, J.C. and Collier, R.J. (1996) Insulinlike growth factors-I–II, somatotropin, prolactin, and placental lactogen are not acute effectors of lipolysis in ruminants. Domestic Animal Endocrinology 13, 239–249. Houseknecht, K.L., Baile, C.A., Matteri, R.L. and Spurlock, M.E. (1998) The biology of leptin: a review. Journal of Animal Science 76, 1405–1420. Jacquemet, N. and Prigge, E.C. (1991) Effect of increased postmilking prolactin concentrations on lactation, plasma metabolites, and pancreatic hormones in lactating goats. Journal of Dairy Science 74, 109–114. Kann, G. (1997) Evidence for a mammogenic role of growth hormone in ewes: effects of growth hormone-releasing factor during artificial induction of lactation. Journal of Animal Science 75, 2541–2549. Kann, G., Carpentier, M.C., Fevre, J., Martinet, J., Maubon, M., Meusnier, C., Paly, J. and Vermeyre, N. (1978) Lactation and prolactin in sheep, role of prolactin in initiation of lactation of milk secretion. In: Robyn, C. and Harter, M. (eds) Progress in Prolactin Physiology and Pathology. Elsevier North-Holland Biomedical Press, Amsterdam/New York, pp. 201–212. Kay, R.N.B. (1985) Seasonal variation of appetite in ruminants. In: Haresign, W. and Cole, D.J.A. (eds) Recent Advances in Animal Nutrition. Butterworths, London, pp. 199–210. Klingenspor, M., Dickopp, A., Heldmaier, G. and Klaus, S. (1996) Short photoperiod reduces leptin gene expression in white and brown adipose tissue of Djungarian hamsters. FEBS Letters 399, 290–294. Knight, C.H. (1993) Prolactin revisited. Hannah Research Institute Yearbook, pp. 72–80.

222

Y. Chilliard and F. Bocquier Knight, C.H. and Flint, D.J. (1995) GH : Prolactin interactions in lactating rodents and ruminants. Hannah Research Institute Yearbook, pp. 73–79. Larsen, T.S., Lagercrantz, H., Riemersma, R.A. and Blix, A.S. (1985a) Seasonal changes in blood lipids, adrenaline, noradrenaline, glucose and insulin in Norwegian reindeer. Acta Physiologica Scandinavica 124, 53–59. Larsen, T.S., Nilsson, N.Ö. and Blix, A.S. (1985b) Seasonal changes in lipogenesis in isolated adipocytes from Svalbard and Norwegian reindeer. Acta Physiologica Scandinavica 123, 53–59. Le Gouic, S., Atgié, C., Viguerie-Bascands, N., Hanoun, N., Larrouy, D., Ambid, L., Raimbault, S., Ricquier, D.P., Gardiola-Lemaitre, B., Pénicaud, L. and Casteilla, L. (1997) Characterisation of a melatonin binding site in Siberian hamster brown adipose tissue. European Journal of Pharmacology 339, 271–278. Lincoln, G.A. and Richardson, M. (1998) Photo-endocrine control of seasonal cycles in body weight, pelage growth and reproduction: lessons from the HPD sheep model. Comparative Biochemistry and Physiology 119C, 283–294. Lincoln, G.A., McNeilly, A.S. and Cameron, C.L. (1978) The effects of a sudden decrease or increase of daylength on prolactin secretion in the ram. Journal of Reproduction and Fertility 52, 305–311. Linzell, J.L. (1973) Innate seasonal oscillations in the rate of milk secretion in goats. Journal of Physiology 230, 225–233. Luo, S., Luo, J., Meier, A.H. and Cincotta, A.H. (1997) Dopaminergic neurotoxin administration to the area of the suprachiasmatic nuclei induces insulin resistance. NeuroReport 8, 3495–3499. Luthman, J. and Johnson, G. (1972) Short term hormonal effects on blood glucose and nonesterified fatty acids in female sheep. Acta Veterinaria Scandinavica 13, 112–127. McElroy, J., Mason, P., Hamilton, J. and Wade, G. (1986) Effects of diets and photoperiod on NE and GDP binding in Siberian hamster brown adipose tissue. American Journal of Physiology 250, R383–388. Malpaux, B., Viguié, C., Thièry, J.C. and Chemineau, P. (1996) Contrôle photopériodique de la reproduction. INRA Productions Animales 9, 9–23. Maule Walker, F.M. and Peaker, M. (1978) Production of oestradiol-17ß by the goat mammary gland during late pregnancy in relation to lactogenesis. Journal of Physiology 284, 71–72. Meier, A.H. and Burns, J.T. (1976) Circadian hormone rhythms in lipid regulation. American Zoology 16, 649–659. Meier, A.H. and Cincotta, A.H. (1996) Circadian rhythms regulate the expression of the thrifty genotype/phenotype. Diabetes Reviews 4, 464–487. Mercer, J.G. (1998) Regulation of appetite and body weight in seasonal mammals. Comparative Biochemistry and Physiology 119C, 295–303. Mercer, J.G., Bruce, C. and Morgan, P.J. (1995) Fat mobilisation in short days is not associated with altered noradrenergic sensitivity of adipocytes in Djungarian hamsters. Physiology and Behaviour 57, 249–253. Mills, S.E., Lemenager, R.P. and Horstman, L.A. (1989) Effect of suppression of prolactin secretion on adipose tissue lipogenesis in the postpartum beef cow. Journal of Animal Science 67, 2904–2912. Milne, J.A., MacRae, J.C., Spence, A.M. and Wilson, S. (1978) A comparison of the voluntary intake and digestion of a range of forages at different times of the year by sheep and the red deer (Cervus Elaphus). British Journal of Nutrition 40, 347–356. Morgan, P.J. and Mercer, J.G. (1994) Control of seasonality by melatonin. Proceedings of the Nutrition Society 53, 483–493. Mrosovsky, N. (1976) Lipid programmes and life strategies in hibernators. American Zoology 16, 685–697.

Direct Effects of Photoperiod on Lipid Metabolism

223

Ortavant, R., Bocquier, F., Pelletier, J.P., Ravault, J.P., Thimonier, J. and Volland-Nail, P. (1988) Seasonality of reproduction in sheep and its control by photoperiod. Australian Journal of Biological Science 41, 69–85. Perier, A., Bocquier, F., Kann, G. and Martinet, J. (1986) Influence de la photopériode et des apports énergétiques pendant la gestation sur les taux plasmatiques de prolactine, de l’hormone de croissance et de l’hormone placentaire lactogène puis sur la production laitière des brebis traite. Reproduction Nutrition Development 26, 391–392. Petitclerc, D. and Zinn, S.A. (1991) Effect of photoperiod on growth and development of ruminants. In: Pearson, A.M. and Dutson, T.R. (eds) Growth Regulation in Farm Animals. Elsevier, London, New York, pp. 427–453. Plaut, K., Bauman, D.E., Agergaard, N. and Akers, R. (1987) Effect of exogenous prolactin administration on lactational performances of dairy cows. Domestic Animal Endocrinology 4, 279–290. Reimers, E., Ringberg, T. and Sorumgaard, R. (1982) Body composition of Svalbard reindeer. Canadian Journal of Zoology 60, 1812–1821. Sheenan, D.B. (1994) Regulation of water and solute transport across mammalian plasma cell membranes by prolactin. Journal of Dairy Research 61, 155–166. Trayhurn, P., Duncan, J.S. and Rayner, D.V. (1995) Acute cold-induced suppression of ob (obese) gene expression in white adipose tissue of mice: mediation by the sympathetic system. Biochemical Journal 311, 729–733. Vernon, R.G. and Finley, E. (1986) Lipolysis in rat adipocytes during recovery from lactation. Biochemical Journal 234, 229–231. Vernon, R.G. and Finley, E. (1988) Roles of insulin and growth hormone in the adaptations of fatty acid synthesis in white adipose tissue during the lactation cycle in sheep. Biochemical Journal 256, 873–878. Vernon, R.G. and Flint, D.J. (1983) Control of fatty acids synthesis in lactation. Proceedings of the Nutrition Society 42, 315–331. Vernon, R.G., Clegg, R.A. and Flint, D.J. (1986) Adipose tissue metabolism in sheep: response to season and its modulation by reproductive state. Hormone and Metabolic Research 18, 308–312. Viguié, C., Thibault, J., Thièry, J.C., Tillet, Y. and Malpaux, B. (1997) Characterization of the short day-induced decrease in median eminence tyrosine hydroxylase activity in the ewe: temporal relationship to the change in luteinizing hormone and prolactin secretion and short day-like effect of melatonin. Endocrinology 138, 499–506. Walker, V.A., Young, B.A. and Walker, B. (1991) Does photoperiod directly influence energy metabolism? In: Wenk, C. and Boessinger, M. (eds) Proceedings of the 12th Symposium on Energy Metabolism of Farm Animals (EAAP Publication no 58), pp. 372–375. Williams, L.M., Lincoln, G.A. and Morgan, P.J. (1995) Melatonin: Sites and mode of action in the sheep. In: Engelhardt, W. von, Leonhard-Marek, S., Breves, G. and Giesecke, D. (eds) Proceedings of the Eighth International Symposium on Ruminant Physiology. Ferdinand Enke Verlag, Stuttgart, pp. 541–557. Youngstrom, T.G. and Bartness, T.J. (1995) Catecholaminergic innervation of white adipose tissue in Siberian hamsters. American Physiological Society 268, R744-R751. Yu, W.H., Kimura, M., Walczewska, A., Karanth, S. and McCann, S.M. (1997) Role of leptin in hypothalamic-pituitary function. Proceedings of the National Academy of Sciences USA 94, 1023–1028.

V

Tissue Growth

13

Muscle Growth and Genetic Regulation

J.J. BASS, M. SHARMA, J. OLDHAM AND R. KAMBADUR AgResearch, Ruakura Agricultural Research Centre, Hamilton, New Zealand

Introduction Growth of a whole animal, or its organs, or tissues, is clearly under a coordinated, integrated control system, of which there is limited understanding (Conlon and Raff, 1999). The size of an organ/tissue depends on the number and size of its cells and extracellular components. Therefore, the growth of a specific tissue, such as muscle, is dependent on the number of myogenic progenitor cells, proliferation of these committed myogenic cells (myoblasts) and the subsequent hypertrophic growth of these muscle fibres. An increase in total muscle mass is the result of a balance between cell proliferation, programmed cell death and cell/fibre growth. Growth of an individual muscle and the overall coordination of muscle growth throughout the body requires the integration of both local and systemic control systems. Research to date has largely focused on local control systems and how they interact with systemic systems, such as the growth hormone (GH) axis. Little work on the coordination of whole-body muscle growth has been undertaken, yet, in numerous selection trials, weight gains in cattle and sheep have shown that coordinated overall growth and associated muscle growth is highly heritable and therefore under genetic control (Notter, 1999). This chapter will briefly review the stages and known control systems of muscle development and then will focus on some of the specific genes which have been shown to regulate muscle growth in ruminants. The continuous development of muscle can be divided into three basic phases for convenience: determination of progenitor cells, proliferation of myoblasts, differentiation and growth of muscle fibres (Fig. 13.1).

Determination Mesodermal cells are the source of all vertebrate skeletal muscle (Cossu et al., 1996) and these originate from progenitor cells in the embryonic epiblast. The mesoderm, which initially consists of lateral plates, divides into somatic mesoderm, splanchnic © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

227

228

J.J. Bass et al.

Possible controlling factors Mesodermal progenitor cells Pax 3 myf5

Determination Myogenic progenic cells Proliferation

MyoD and myf5 Myoblasts

Proliferation differentiation

MyoD and myf5 Terminally differentiated myoblasts

Myostatin Myogenin

Fusion

IGF-II

Myotubes Maturation MRF4 Mature muscle fibres Hypertrophy

IGF-I

IGF-I

Fig. 13.1. Determination, proliferation, fusion and differentiation of myogenic cells and controlling factors.

mesoderm, intermediary mesoderm and paraxial mesoderm. The paraxial mesoderm then develops into somites (Hauschka, 1994). The ventral part of the somite mesoderm forms the sclerotome (Fig. 13.2) and the dorsal part becomes the dermomyotome that develops into dermatome and myotome. Both the axial and appendicular muscles are derived from myotome. The dorsal myotome forms the epiaxial back muscles, while cells from the lateral myotome migrate to the limbs to form limb muscles or hypoaxial muscles of the ventral body wall. The fate of mesodermal cells appears to depend on external factors from surrounding tissue. The position of mesoderm relative to the neural tube, notochord, dorsal ectoderm and lateral plate mesoderm plays an important role in determining whether somatic cells will become myogenic precursor cells. A number of factors from these surrounding tissues have been identified and associated with the induction of

229

Muscle Growth and Genetic Regulation

Medial myf5

Dermomyotome Pax 3

Neural tube

Lateral c-met

Myotome

Notochord Sclerotome Pax1

Fig. 13.2. Induction of myogenesis by factors from surrounding tissues.

myogenic cells (Cossu et al., 1996; Capdevila et al., 1998) (Fig. 13.1). Examples of inducing factors are: Pax 3, which has been shown to be essential for the determination of myogenic progenitor cells and their migration into limb muscles (Cossu et al., 1996) and sonic hedgehog, which has been identified as a candidate for inducing the onset of myogenesis (Marcelle et al., 1997). The myogenic regulatory factors (MRFs), members of the helix-loop-helix group of transcription factors, are expressed in the medial dermomyotome and may have a role in determination of myogenic cells as well as late stages of muscle fibre development. Both Myf5 and MyoD are believed to be essential for the early development of muscle, as mice deficient in both lack normal myoblasts and skeletal muscle (Edmondson and Olson, 1993). An understanding of the mechanisms that control the formation of myogenic progenitor cells will allow the future identification of critical gene pathways for ruminants as well as other species. At the moment, no observed variations in these early myogenic pathways have been associated with changes in muscle mass of ruminants.

Proliferation The second stage of muscle development is proliferation. During this stage, committed, mononucleated myogenic cells (myoblasts) first proliferate and then differentiate into fusion-competent myoblasts. These fusion-competent myoblasts divide and eventually fuse end-to-end, forming long, multinucleated myotubes. Myoblasts can be divided into three types on developmental age and morphology in culture (Franzini et al., 1994). These are the embryonic or early myoblasts, isolated from embryonic muscles prior to and during primary myotube formation; the fetal or late myoblasts, isolated from fetal muscles throughout secondary fibre formation; and the satellite cells,

230

J.J. Bass et al.

isolated from adult muscles. The number of myoblast divisions may be important in regulating the skeletal musculature in the growing animal. Myoblasts have limited capacity to proliferate (Grounds and Yablonka-Reuveni, 1993), compared with other cells. Whether a myoblast continues dividing or differentiates appears to be dependent on extracellular cues, such as growth factors, extracellular matrix and cell-to-cell interactions. A large number of in vitro studies have indicated that myoblast proliferation can be regulated by growth factors (Florini and Magri, 1989). To date, the major growth factors which affect myoblast proliferation are fibroblast growth factor, insulin, insulin-like growth factors (IGF), platelet-derived growth factor and transforming growth factor-␤ (TGF-␤). Other mitogens such as epidermal growth factor and leukaemia inhibiting factor have also been shown to affect myoblast proliferation (Hauschka, 1994). Growth factors often have different biological effects which appear to depend on their concentration and the developmental stage of the myoblasts. This divergence of function is seen with IGF-I that can stimulate proliferation as well as maintain differentiation (reviewed by Florini et al., 1996) and in TGF-␤ that inhibits both proliferation and differentiation of myogenesis (Massagué et al., 1986). An example of how a change in the control system of myoblast proliferation can affect muscling of cattle is to be found in the ‘double-muscled’ Belgian Blue breed. In certain breeds, intense selection over many generations has resulted in extreme muscle hypertrophy known variously as double-muscling, culard, etc. (Boccard, 1981). This type of muscle hypertrophy is mainly associated with an increase in the number of muscle fibres, especially secondary muscle fibres. It has been postulated that the increased fibre number results from increased numbers of late myoblasts, either because of increased myoblast proliferation or because of delayed differentiation into myotubes. The inheritance of double-muscling in Belgian Blue cattle has been identified as a monogenic autosomal segregation pattern (Hanset and Michaux, 1985; Charlier et al., 1995). The muscular hypertrophy (mh) locus has been termed ‘partially recessive’, because, while a single copy of the allele may have some effect, both are required for full expression of the double-muscled phenotype. Gene mapping (Charlier et al., 1995) of the Belgian Blue cattle localized the mh gene to the centromeric end of the bovine chromosome 2 (BTA2) linkage group. The mh locus, in more refined studies in the Piedmontese breed, localized the gene to a 3–5-cm interval near the centromere of BTA2 close to the position of the ␣-collagen type III (COL3A1) locus (Casas et al., 1999). Recently, growth differentiation factor-8 (GDF-8) (McPherron et al., 1997), a member of the TGF-␤ superfamily, was disrupted in mice. These GDF-8 null mice were significantly larger than their wild type counterparts, and showed an increase in body weight resulting from a two- to threefold increase in muscle mass. This increase in muscle was due to an increase in the number of muscle fibres. As GDF-8 seemed to function as an inhibitor of muscle growth, it was renamed myostatin. The myostatin gene has since been mapped to the same interval as the mh locus in cattle by genetic linkage (Smith et al., 1997). The finding that the myostatin null mice have a similar phenotype to double-muscled cattle and that the gene maps to a similar site, suggests that the myostatin and the muscular hypertrophy genes are one and the same. This hypothesis was further supported by finding an 11 bp mutation in the myostatin gene in Belgian Blue cattle and a point mutation in Piedmontese cattle, both doublemuscled breeds (Kambadur et al., 1997).

Muscle Growth and Genetic Regulation

231

More recently, Grobet et al. (1997) have identified seven DNA sequence polymorphisms, of which five were predicted to disrupt the function of myostatin. These studies clearly demonstrate that the double-muscling phenotype in cattle is genetically heterogeneous and involves several mutations in the myostatin gene. The 11 bp deletion in the Belgian Blue cattle results in the loss of three amino acids, that causes a frame-shift after amino acid 273. This frame-shift leads to a stop codon after amino acid 286 that is predicted to produce a truncated, biologically inactive protein. In the Piedmontese breed, a mutation at position 941 bp results in the loss of a cysteine at amino acid 314 (Kambadur et al., 1997). This cysteine has been shown to be essential in other TGF-␤ family members in order to form a cysteine knot, which stabilizes the TGF-␤ dimer. In general, these studies suggest that myostatin is probably the mh locus and that it acts as an inhibitor of muscle development by limiting muscle fibre number, and to some extent, muscle fibre size. When the development of proliferating and fusing myoblasts was followed, it was found that myostatin mRNA was elevated during primary fibre formation through to the early stages of secondary fibre formation. This suggests that myostatin may participate in the proliferation and terminal differentiation of late myoblasts (Oldham et al., 1998). This study also suggested that a link exists between MyoD and myostatin, as MyoD expression was increased during the fibre formation in muscle from doublemuscled cattle, which lack myostatin. Interestingly, the different myostatin mutations found in cattle do not give similar increases in muscle mass. This suggests that the myostatin proprotein and the mutant proteins inhibit muscle growth differently. The possibility also exists that associated muscle-controlling genes have been differentially selected in the different breeds. A further level of control of muscle development by myostatin has been identified in the compact hypermuscular mouse (Szabo et al., 1998). The mutation that causes an increase in muscle mass in the compact mouse is a deletion in the pro-peptide region, which precedes the proteolytic processing site of myostatin. The pro-peptide region, by analogy with TGF-␤ (Miyazono et al., 1991), may be involved in the folding and secretion of myostatin. Such changes possibly decrease the biological activity of myostatin, but do not completely remove it. Myostatin appears to be a tissue-specific inhibitor of myoblast proliferation that can have a direct effect on myoblast proliferation and so also on muscle fibre number. Examples of factors that stimulate rather than inhibit myoblast proliferation include IGF-I and -II, as both have been shown to enhance myoblast proliferation in culture (Florini et al., 1996). Whether both IGFs are local growth factors that are critical in controlling myoblast proliferation in vivo is unclear, because local expression of IGF-II increases at the end of secondary myofibre formation, whereas IGF-I does not show marked changes associated with stages of fetal muscle formation in cattle (Oldham et al., 1998). In most species, serum concentrations of IGF-I, but not IGF-II, correlate positively with fetal weight (Gluckman and Brinsmead, 1976; Humbel, 1990). The IGFs have been related to fetal growth in general. When gene targeting is used to disrupt IGF-II, the offspring are only 30% of the weight of the wild type (Baker et al., 1993), while IGF-I null mutants are growth retarded, depending on the magnitude of the reduction in IGF-I concentration (Liu et al., 1998). While IGF-I appears to act locally on developing muscle (Tollefsen et al., 1989), it dramatically increases IGF-II mRNA in regenerating adult muscle when satellite cells from healthy muscle divide and

232

J.J. Bass et al.

migrate into damaged areas to initiate regeneration (Kou and Rotwein, 1993; Kirk et al., 1996). This indicates that IGF-II is important in myoblast proliferation and differentiation. GH, which is the hormone controlling the expression of IGF-I, does not seem to be directly involved in myoblast proliferation, as mice overexpressing GH do not necessarily show increased muscle growth or increased IGF-I expression at birth (Liu et al., 1998). However, the treatment of pregnant sows with GH early in gestation increases both the number of muscle fibres and the liveweight at birth of their offspring (Rehfeldt et al., 1993). Such an effect of GH may be indirect via maternal changes rather than acting directly on the fetus.

Differentiation When myoblasts leave the cell cycle and become terminally differentiated, they fuse with one another to form myotubes (Fig. 13.1). Myotubes are immature muscle fibres which form in distinct phases. Primary myotubes are thought to arise from end-to-end fusion of early myoblasts. Late myoblasts then accumulate on the surface of the preexisting primary myotubes and fuse to form secondary myofibres. In all muscles, there are many more secondary myotubes than primary myotubes. In large animals such as sheep and cattle, there is possibly a third phase of myogenesis, where the tertiary myofibres use the secondary fibres as a scaffolding, similar to the way developing secondary fibres use primary fibres for support. These tertiary fibres have a different myosin expression and are mainly distributed along the border of muscle fascicles (Franzini et al., 1994). Any change in the number of myofibres which form could have a profound effect on the total muscle mass of the mature animal. A number of the factors that regulate expression of muscle-specific genes following commitment to terminal differentiation have been established. The MRFs are the best characterized of these factors (Edmondson and Olson, 1993) and their sequential activation commits cells to induce genes required to establish terminally differentiated muscle cells. Of the MRFs, MyoD and Myf-5 are involved in the regulation of myoblasts and satellite cell proliferation, whereas myogenin induces differentiation (Olsen and Klein, 1994). MRF4 is expressed late in muscle differentiation and may share some of the functions of myogenin (Buckingham, 1994). MRFs, however, do not function alone, they require transcriptional co-activators such as the myocyte enhancer factor-2 (MEF2) before they can regulate myoblast and satellite cell proliferation (Olson et al., 1995). MRFs are also expressed in adult muscle while MyoD protein accumulates in satellite cells of regenerating myotubes and skeletal muscle fibres. This appears to be neurally regulated (Koishi et al., 1995). There are now indications that polymorphisms in the MRFs could be used as selection markers for improved growth rates in pigs. Polymorphic sites at the myogenin and Myf-5 gene loci have been associated with increases in birth weight, postnatal growth and weight of lean meat in pigs (Te Pas et al., 1998). However, Myf-5 polymorphisms were not related to increased growth rates. Furthermore, expression levels of MyoD showed no relation to selection for liveweight gain (Te Pas et al., 1998). Similar associations have not so far been reported for ruminants.

Muscle Growth and Genetic Regulation

233

Terminal muscle cell differentiation eventually results in the sequential expression of muscle-specific genes, which include desmin, acetylcholine receptors and specific isoforms of actin, myosin heavy chain, troponin, tropomyosin and creatine kinase (Schiaffino and Reggiani, 1996). The expression of these and other genes leads to a differentiated muscle that is fully functional. Polymorphic variations have been reported for some of these genes and many, like myogenin, could be associated with improved meat production. An example is the myosin light chain-1-fast gene that modulates the interactions between actin and myosin during contraction. This gene undergoes differential splicing, which results in two products in chickens (Rushbrook and Somes, 1985) and in cattle (A. Clark, personal communication). However, as yet, no association has been found between these polymorphisms and any desirable muscle characteristics in cattle.

Hypertrophy The bulk of muscle growth results from satellite cell division and incorporation of daughter cells into the multinucleated fibre. This enables the nuclei/cytoplasmic ratios of the muscle fibre to be maintained, even though the nuclei of the fibre are unable to undergo division. During normal growth, and in response to stretch, muscle fibres lengthen by the addition of new sarcomeres at the ends of these fibres (Williams and Goldspink, 1973). Under conditions of normal growth, the longitudinal and radial growth of muscle fibres is highly correlated. However, in sheep heavier than 45 kg, there is little further increase in radial growth of muscle fibres, while in cattle, radial fibre growth is rapid at birth. This growth, as in sheep, does level off in older animals (Swatland, 1984). Specific muscle hypertrophy can be work-induced, which results in the increased local expression of IGF-I and an increase in insulin sensitivity in association with increased glucose metabolism, amino acid transport, protein synthesis and protein degradation. More recently, the expression of the muscle growth inhibitor, myostatin, has been shown to decrease immediately after stretch-induced hypertrophy is initiated (J. Martyn et al., personal communication). However, while muscle fibre hypertrophy is found in the myostatin knockout mouse (McPherron et al., 1997), it has not been identified in double-muscled cattle that have a mutated, non-active myostatin (Boccard, 1981). There are a number of animal models that show an abnormal increase in postnatal muscle size. The callipyge sheep was derived from a ram with extreme muscling, especially in the hindquarter (Cockell et al., 1994). The fine mapping of the CLPE locus to ovine chromosome 18 has so far failed to identify a candidate gene for this trait. The compact mouse (Szabo et al., 1998) also shows a postnatal increase in fibre size, resulting in enlarged hindquarter muscles. This muscle hypertrophy has been associated with a mutation in the myostatin pro-peptide region, which precedes the proteolytic processing site (Szabo et al., 1998). This mutation may, as in the TGF-␤ proteins (Miyazono et al., 1991), affect the folding of myostatin and hence its biological activity. The lack of information concerning muscle hypertrophy compared with the number of publications on myoblast proliferation and differentiation results from a lack of suitable in vitro models, in which satellite cell division may be observed in association with increased muscle fibre growth. In future, the clever use of transgenic models and

234

J.J. Bass et al.

Cre-Lox knockouts (Sauer, 1998), which may be both tissue- and time-based, will allow the identification of genes which may be controlling postnatal growth of muscle fibres, the period during which muscle size increases the most.

Acknowledgements Research into growth physiology by AgResearch is funded by the New Zealand Foundation of Research Science and Technology. We thank the organizers for inviting this review and the colleagues who have contributed.

References Baker, J., Liu, J-P., Robertson, E.J. and Efstratiadis, A. (1993) Role of insulin-like growth factors in embryonic and post-natal growth. Cell 75, 73–82. Boccard, R. (1981) Facts and reflections on muscular hypertrophy in cattle: double muscling or culard. In: Ralson, L. (ed.) Developments in Meat Science, 2. Applied Science Publishers, pp. 1–28. Buckingham, M. (1994) Which myogenic factors make muscle? Current Biology 4(1), 61–63. Capdevila, J., Tobin, C. and Johnson, R.L. (1998). Control of dorsoventral somite patterning by Wnt-1 and ␤ catenin. Developmental Biology 193, 182–194. Casas, E., Shackelford, S.D., Keele, J.W., Stone, R.T., Kappes, S.M. and Koohmaraie, M. (2000) Quantitative trait loci affecting growth and carcass composition of cattle segregating alternate forms of myostatin. Journal of Animal Science 78(3), 560–569. Charlier, C., Coppieters, W., Farnir, F., Grobet, L., Leroy, P.L., Michaux, C., Mni, M., Schwers, A., Vammanshoven, P., Hanset, R. and Georges, M. (1995) The mh gene causing doublemuscling in cattle maps to bovine chromosome2. Mammalian Genome 6, 788–792. Cockell, N.E., Jackson, S.P., Shay, T.L., Nielsen, D., Moore, S.S., Steele, M.R., Barendse, W., Green, R.D. and Georges, M. (1994) Chromosomal localization of the callipyge gene in sheep (Ovis aries) using bovine DNA markers. Proceedings of the National Academy of Sciences USA 91, 3019–3023. Conlon, I. and Raff, M. (1999) Size control in animal development. Cell 96, 244–245. Cossu, G., Tajbakhsh, S. and Buckingham, M. (1996) How is myogenesis initiated in the embryo. Trends in Genetics 12(6), 218–223. Edmondson, D.G. and Olson, E.N. (1993) Helix-Loop-Helix proteins as regulators of musclespecific transcription. The Journal of Biological Chemistry 268(2), 755–758. Florini, J.R. and Magri, K.A. (1989). Effects of growth factors on myogenic differentiation. American Journal of Physiology 256 (Cell Physiol 25), C701–C711. Florini, J.R., Ewton, D.Z. and Coolican, S.A. (1996) Growth hormone and the insulin-like growth factor system in myogenesis. Endocrine Reviews 17, 481–517. Franzini, T., Armstrong, C. and Fischman, D.A. (1994) Morphogenesis of skeletal muscle fibres. In: Engel, A.G. and Franzini-Armstrong, C. (eds) Myology pp.74–96. Gluckman, P.D. and Brinsmead, M.W. (1976) Somatomedin in cord blood: relationship to gestational age and birth size. Journal of Clinical Endocrinology Metabolism 43(6), 1378–1381. Grobet, L., Royo Margin, L.J., Poncelet, D., Pirottini, D., Brouwers, B., Riquet, J., Schoeberlain, A., Punner, S., Menissier, F., Massabunda, J., Fries, R., Hanset, R. and George, M. (1997) A deletion in the myostatin gene causes double muscling in cattle. Nature Genetics 17, 71–73. Grounds, M.D. and Yablonka-Reuveni, Z. (1993) Molecular and cell biology of skeletal muscle

Muscle Growth and Genetic Regulation

235

regeneration. In: Partridge, T.A. (ed.) Molecular and Cell Biology of Muscular Dystrophy, Chapman & Hall, New York, pp. 210–256. Hanset, R. and Michaux, G. (1985) On the genetic determination of muscular hypertrophy in the Belgian White and Blue cattle breed. I. Experimental data. Genetique, Selection, Evolution 17, 354. Hauschka, S. (1994) The embryonic origin of muscle. In: Engel, A.G. and Franzini-Armstrong, C. (eds) Myology, Vol. 1. pp. 3–73. Humbel, R.E. (1990) Insulin-like growth factors I and II. European Journal of Biochemistry 190(3), 445–462. Kambadur, R., Sharma, M., Smith, T.P.L. and Bass, J.J. (1997) Mutations in myostatin (GDF8) in double-muscled Belgian Blue and Piedmontese cattle. Genome Research 7, 910–916. Kirk, S.P., Whittle, M.A., Oldham, J.M., Dobbie, P.M. and Bass, J.J. (1996) GH regulation of the type 2 IGF receptor in regenerating skeletal muscle of rats. Journal of Endocrinology 149, 81–91. Koishi, K., Zhang, M., McLennan, I. and Harris, A.J. (1995) MyoD protein accumulates in satellite cells and is neurally regulated in regenerating myotubes and skeletal muscle fibers. Developmental Dynamics 202(3), 244–254. Kou Kou, and Rotwein, P. (1993) Transcriptional activation of the insulin-like growth factor-II gene during myoblast differentiation. Molecular Endocrinology 7(2), 291–302. Liu, J.I., Grinbery, A., Westphal, H., Sauer, B., Accili, D., Karas, M. and Le Roith, D. (1998) Insulin-like growth factor-I affect perinatal lethality and postnatal development in a gene dosage-dependent manner: manipulation and using the Cre/LoxP system in transgenic mice. Molecular Endocrinology 12(9), 1452–1462. McPherron, A.C., Lawler, A.M. and Lee, S-J. (1997) Regulations of skeletal muscle mass in mice by a new TGF␤ superfamily member. Nature 387, 87–90. Marcelle, C., Stark, M.R. and Bronner-Fraser, M. (1997) Co-ordinate actions of BMPs, Wnts, shh and noggin mediate patterning of the dorsal somite. Development 124, 3955–3963. Massagué, J., Cheifetz, S., Endo, T. and Nadal-Ginard, B. (1986). Type ␤ transforming growth factor is an inhibitor of myogenic differentiation. Proceedings of the National Academy of Sciences USA 83, 8206–8210. Miyazono, K., Olofosson, A., Colosetti, P. and Heldin, C.H. (1991) A role of the latent TGF␤1-binding protein in the assembly and secretion of TGF-␤1. European Molecular Biology Organisation Journal 10, 1091–1101. Notter, D.R. (1999) The importance of genetic diversity in livestock populations of the future. Journal of Animal Science 77, 61–69. Oldham, J., Sharma, M., Martyn, J., Kambadur, R. and Bass, J. (1998) Myostatin and myogenic regulatory factors (MRFs), but not insulin-like growth factors (IGFs), differ between normal and double muscled cattle. Proceedings of the 80th Conference of the Endocrine Society, pp. 3–310. Olson, E.N. and Klein, W.H. (1994) bHLH factors in muscle development. Genes and Development 8, 1–8. Olson, E.N., Perry, M. and Schulz, R.A. (1995) Regulation of muscle differentiation by the MEF2 family of MADS box transcription factors. Developmental Biology 172, 2–14. Rehfeldt, C., Fiedler, I., Weikard, R., Kanitz, E. and Ender, K. (1993) It is possible to increase skeletal muscle fibre number in utero. Bioscience Reports 13(4), 213–220. Rushbrook, J.I. and Somes, R.G. (1985) Myosin light chain-1: Genetic analysis of three variants found in fast white chicken muscle and investigation of linkage with the muscular dystrophy gene. Biochemical Genetics 23(1/2), 17–27. Sauer, B. (1998) Inducible gene targeting in mice using the Cre/Lox System. Methods. 14(4), 381–392.

236

J.J. Bass et al. Schiaffino, S. and Reggiani, C. (1996) Molecular diversity of myofibrillar proteins: Gene regulation and functional significance. Physiological Reviews 76(2), 371–423. Smith, T.P.L., Lopez-Corrales, N.L., Kappes, S.M. and Sonstegaard, T.S. (1997) Myostatin maps to the interval containing the bovine mh locus. Mammalian Genome 8, 742–744. Swatland, H.J. (1984) (ed.) Structure and Development of Meat Animals. Prentice-Hall, Englewood Cliffs, New Jersey, pp. 201–254. Szabo, G., Dallmann, G., Muller, G., Patthy, L., Soller, M. and Varga, L. (1998) A deletion in the myostatin gene causes the compact (Cmpt) hypermuscular mutation in mice. Mammalian Genome 9, 671–672. Te Pas, M.F.W., Harders, F.H., Soumillion, A., Verburg, F.J. and Meuwissen, T.H.E. (1998) The MyoD gene family influences birth weight, postnatal, growth rate and carcass meat content of pigs. Proceedings of the XXVI International Conference on Animal Genetics, E048. Tollefsen, S.E., Lajara, R., McCusker, R.H., Clemmons, D.R. and Rotwein, P. (1989) Insulinlike growth factors (IGF) in muscle development. The Journal of Biological Chemistry 264(23), 13810–13817. Williams, P. and Goldspink, G. (1973) The effect of immobilisation on the longitudinal growth of striated muscle fibres. Journal of Anatomy 116, 45–55.

14

Control and Manipulation of Hyperplasia and Hypertrophy in Muscle Tissue

P.J. BUTTERY, J.M. BRAMELD AND J.M. DAWSON Division of Nutritional Biochemistry, School of Biological Sciences, University of Nottingham, Loughborough, UK

Introduction – skeletal muscle growth The growth of skeletal muscle involves increases in both cell number (hyperplasia) and cell size (hypertrophy). Hyperplasia involves the mononuclear skeletal muscle precursor cells (myoblasts and satellite cells), which subsequently become post-mitotic, align and fuse (differentiate) to form the multinuclear muscle fibres. Muscle fibres are then able to increase in cell size (hypertrophy) via the accretion of more protein. The total number of fibres in a muscle appears to be fixed at, or shortly after birth, with postnatal growth of muscle being entirely due to elongation and widening of the existing muscle fibres. However, this does not mean that the muscle precursor cells cease to grow at this stage, as the DNA content continues to increase until the animal approaches its mature size. Extra nuclei are recruited into muscle fibres from mononuclear myogenic cells known as satellite cells, which are enclosed beneath the basement membrane of the muscle fibres (see Harper and Buttery, 1992). Muscle cell differentiation therefore takes place during the mid–late gestation and neonatal periods, with differentiation in the neonatal period involving the fusion of satellite cells with the existing muscle fibres.

Myogenic cell determination and differentiation Whether a mesodermal cell becomes a skeletal muscle cell or some other cell type is determined at a very early stage of embryo development. A family of nuclear proteins has been identified which function as activators of muscle-specific gene transcription (i.e. transcription factors), and forced expression of any one of them can convert fibroblasts to muscle cells. They are called the MyoD family and are discussed in detail in Chapter 13.

© CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

237

238

P.J. Buttery et al.

Control of hyperplasia and differentiation Effects of growth factors, hormones and nutrients in vitro A wide variety of growth factors and metabolic hormones have been shown to affect skeletal muscle cell proliferation and/or differentiation in vitro (Table 14.1). Insulinlike growth factors (IGFs), epidermal growth factor (EGF) and insulin consistently stimulate both proliferation and differentiation, the fibroblast growth factors (FGFs) stimulate proliferation and inhibit differentiation, and the transforming growth factor␤ (TGF-␤) family tend to inhibit both proliferation and differentiation. Differentiation of muscle cells in culture can be induced by the reduction of the serum content of the culture medium. The mechanism for this is thought to be via a reduction in the blood-borne inhibitors of differentiation, together with increases in locally produced activators, such as the IGFs. Differentiation has also been shown to be dependent upon other components of the cell culture medium, including the type of serum used (Doumit and Merkel, 1992), the type of medium and the substratum on which the cells are grown (Dodson et al., 1990). Studies investigating the manipulation of specific nutrients in the media are lacking. Published studies of direct nutritional effects on muscle cells in culture mainly involve micronutrients. A lack of zinc inhibits C2C12 myoblast differentiation and decreases expression of MyoD and myogenin mRNA (Petrie et al., 1996). Similarly, a lack of calcium can also inhibit muscle cell fusion (Morris et al., 1976). An increase in m-calpain, which requires calcium for activity, is seen during differentiation of fetal chicken myoblasts (Kwak et al., 1993). The increase in m-calpain correlates with the elevated cleavage of filamin which occurs during the fusion process, suggesting that mcalpain plays an important role in the cytoskeletal reorganization that takes place during myoblast fusion. Addition of specific inhibitors for proteosome or calpains to the culture medium has also been shown to inhibit muscle differentiation (Ueda et al., 1998), while addition of IGF-I, which stimulates differentiation, increases m-calpain mRNA concentrations (Hong and Forsberg, 1994) in rat L8 myotubes. Retinoic acid induces myogenic differentiation and myogenin synthesis in a rat rhabdomyosarcoma cell line (Arnold et al., 1992). The effects of retinoids are mediated through two receptor subtypes, namely the retinoic acid receptors (RARs) and the retinoid X receptors (RXRs). Expression of RAR mRNA has been shown to be repressed during C2C12 myoblast differentiation, while that of RXR mRNA was induced (Downes et al., 1994). The unsaturated fatty acid, linoleic acid, has also been shown to stimulate myogenic differentiation in rat satellite cells (Allen et al., 1985), but the mechanism for this effect is not clear. Dodson et al. (1990) suggested that differentiation is increased in low-glucose Dulbecco’s modified Eagle’s medium (DMEM), compared with high-glucose DMEM. However, we (Brameld et al., 1999b) have found no effect of glucose concentration on differentiation of primary sheep muscle cells in culture, as measured by changes in intracellular creatine kinase activity, nor on the expression of IGF-I, IGF-II or growth hormone receptor (GHR) genes. Nutrients could affect myoblast differentiation via effects on locally produced growth factors or the hormones known to be influenced by diet (e.g. insulin and IGF-I).

Factor Proliferation Differentiation

Insulin

IGF

GH

FGF

EGF/ TGF-␣

HGF

≠ ≠

≠ ≠

Æ ≠Æ

≠Æ Ø

≠Æ ≠

≠ Ø

TGF-␤ PDGF Dex ØÆ Ø

≠Æ Ø

≠Æ ÆØ

T3 Æ ≠Æ

Test ␤-agonist Æ Ø

≠Æ ≠Æ

LIF

IL-6

RA

LA

CGRP

≠ ND

≠ ND

ND ≠

ND ≠

ND ≠

IGF, insulin-like growth factors -I and -II; GH, growth hormone; FGF, fibroblast growth factor; EGF/TGF␣, epidermal growth factor/transforming growth factor-␣ (same receptor); HGF, hepatocyte growth factor; TGF-␤, transforming growth factor-␤; PDGF, platelet-derived growth factor; Dex, Dexamethasone; T3, tri-iodothyronine; Test, testosterone; ␤-agonist, clenbuterol, ractopamine or isoproterenol; LIF, leukaemia inhibitory factor; IL-6, interleukin-6; RA, retinoic acid; LA, linoleic acid; CGRP, calcitonin gene related peptide; ≠, stimulates; Ø, inhibits; Æ, no effect; ND, no data found.

Control and Manipulation of Hyperplasia and Hypertrophy

Table 14.1. Effects of various growth factors and hormones on muscle precursor cell proliferation and differentiation (compiled from Allen et al., 1985; Arnold et al., 1992; Brameld et al., 1998; Gal-Levi et al., 1998).

239

240

P.J. Buttery et al.

Genetic effects on hyperplasia and muscle fibre number The double-muscled (DM) syndrome, a hereditary condition in which cattle possess 40% more fibres than normal cattle at birth, is thought to be caused by an increase in muscle cell hyperplasia prior to differentiation. Increased growth factor activity during early fetal development has been detected in the serum of DM fetuses (Gerrard and Judge, 1993); myoblasts from DM fetuses proliferate more rapidly than those from normal fetuses (Quinn et al., 1990). Gerrard and Grant (1994) suggested that maximal local IGF-II mRNA expression is delayed in DM fetuses. This may relate to the stimulatory effects of IGF-II on muscle differentiation (i.e. differentiation is also delayed in DM fetuses, enabling greater proliferation to occur). A similar conclusion was recently drawn from studies of the high-growth (hg) locus in mice (Summers and Medrano, 1997). The 40% increase in their adult body weight and the proportional increase in their skeletal muscle weight was associated with delayed fusion and prolonged proliferation of high-growth myoblasts. This resulted in muscle fibre hyperplasia at birth, with only moderate muscle fibre hypertrophy. However, other factors are probably also involved. Recent studies have demonstrated a number of mutations in the myostatin gene in DM cattle (see Bass et al., Chapter 13, this volume). Myostatin (also called GDF-8) was originally identified in mice (McPherron et al., 1997), and was shown to be a member of the TGF-␤ family and therefore an inhibitor of skeletal muscle cell hyperplasia and differentiation. Gene targeting to disrupt the GDF-8 gene in mice increased muscle mass by 2–3 times by a combination of hyperplasia and hypertrophy. The identification of other growth factors involved in double-muscling (be they DMspecific factors or altered concentrations of typical growth factors at critical stages of growth) may facilitate the increase of muscle fibre number in other species. For example, FGFs are expressed by muscle cells in culture (Moore et al., 1991). They stimulate muscle cell proliferation (Table 14.1), but inhibit differentiation, myogenin gene expression (Brunetti and Goldfine, 1990) and also IGF-II gene expression (Rosenthal et al., 1991), and could therefore play a role in controlling the timing of both proliferation and differentiation. There is a strong positive correlation between muscle fibre number and postnatal growth. Fast-growing strains of animals have more muscle fibres than slower-growing strains (Ezekwe and Martin, 1975; Miller et al., 1975). This correlation is apparent even within the same strain of animals; pigs between 50 and 80 kg body weight show a positive correlation between muscle fibre number and average daily gain (Dwyer et al., 1993). In younger pigs (up to 25 kg body weight), however, growth rate correlates with birth weight and not muscle fibre number. Comparison of IGF-I gene expression in three breeds of pig with differing carcass characteristics and growth rates demonstrated significant differences in skeletal muscle (and adipose tissue) IGF-I expression (Brameld et al., 1996a). IGF-I expression in longissimus dorsi (LD) muscle (and adipose tissue) was lower in the slowest growing Meishan ⫻ Landrace than in the other two breeds, with no difference in hepatic IGF-I expression. Thus, local muscle IGF-I expression and its subsequent effects on muscle growth may account for differences in growth rate.

Control and Manipulation of Hyperplasia and Hypertrophy

241

Effects of prenatal nutrition/hormones (fetal programming) The supply of maternal nutrients can influence both the development of the fetus in utero and in the early neonatal period, when the young is dependent on mother’s milk. Nutritional supply in utero is believed to induce programming of the fetal phenotype, with lifelong consequences. Variation in nutrient supply, particularly nutritional restriction, may have temporary effects on development from which the animal can fully recover when nutrient supply is restored. The effects may also be permanent and have long-term consequences for postnatal growth and body composition. In particular, fetal undernutrition at critical stages of development is believed to cause reduced or disproportionate fetal growth (see Desai and Hales, 1997). Nutrition may influence fetal growth at all stages of development, from preimplantation to parturition. The effects of nutrient modulation on fetal development depends upon the timing, duration and severity of the nutritional insult (see Robinson et al., 1999), but the response may differ between species (Hegarty and Allen, 1978). In utero undernutrition may be caused either by maternal feed restriction or through maternal constraint in supplying nutrients to the developing fetus, such as impaired uterine blood supply or placental development, which in turn may be influenced by maternal nutrition (see Robinson et al., 1999). Nutrient restriction in early gestation can cause a permanent reduction in muscle fibre number. Muscle fibre formation begins in early development and is completed in utero in precocious species such as sheep, pigs and guinea pigs. In these species, fetal undernutrition can lead to runted animals with fewer muscle fibres (Powell and Aberle, 1981; Dwyer and Stickland, 1994). Fibre number is fixed soon after birth in all mammals and postnatal growth involves only hypertrophy of the existing fibres. Muscle fibre number is thus a critical determinant of muscle mass. Animals born with fewer muscle fibres remain smaller throughout postnatal life and never reach the same mature size as their well-nourished siblings, although fibre diameter may be larger in these animals (Powell and Aberle, 1981). Primary muscle fibres develop relatively early in gestation when the fetus is making little nutritional demand on the mother. Their development appears to be genetically determined and insensitive to environmental influences such as nutrition (Handel and Stickland, 1987). Primary muscle fibres generally develop into type I (slow-twitch) fibres with an oxidative metabolism. Muscles with a predominance of slow-twitch fibres tend to be less affected by undernutrition due to the lower proportion of secondary fibres in these muscles (Dwyer et al.,1995). Secondary muscle fibres, which form around the primary fibres and constitute the major fibre population in developed muscle, are reduced in nutritionally-restricted animals (Dwyer et al., 1995) as a result of reduced secondary myoblast proliferation. This may be due to reduced placental size and impaired nutrient transfer (Dwyer et al., 1992). The reduction in secondary fibre number was prevented in maternal feedrestricted guinea pigs by replenishing either the protein or carbohydrate components of the diet, indicating that fibre number is reduced only when maternal diets are deficient in both protein and carbohydrate; fat supplementation had no effect (Dwyer and Stickland, 1994). This suggests that the effect is mediated by the carbohydrate, presumably glucose. Doubling feed intake of pregnant sows in the period immediately before muscle fibre hyperplasia (25–50 days of gestation), resulted in an increased mean secondary muscle fibre number in the litter (Dwyer et al., 1994). An increase in

242

P.J. Buttery et al.

both primary and secondary fibre size and in the secondary:primary fibre ratio is seen in muscles from large sheep fetuses obtained from in vitro embryo culture (Maxfield et al., 1998). In vitro embryo culture in serum-supplemented media appears to significantly stimulate growth in some fetuses and this growth advantage is retained throughout gestation (Sinclair et al., 1998). McCoard et al. (1997) demonstrated similar reductions in muscle fibre number in constrained ovine fetuses at 140 days gestation, with the constraints being the presence of twin or single fetuses, or fetuses due to be born in the spring or autumn. Both these constraints result in reduced fetal growth, presumably due to reduced availability of nutrients per fetus. The previously described relationships between muscle fibre number and both postnatal growth rates and gain:feed ratios (Dwyer et al., 1993) demonstrate the importance of maternal nutrition during pregnancy on subsequent growth potential. In species such as rats, a period of in utero undernutrition in early gestation may not result in permanent reduction in muscle fibre numbers if adequate nutritional realimentation occurs in the remaining gestation and early postnatal life (Hegarty and Allen, 1978; Wilson et al., 1988); secondary fibre proliferation continues into early postnatal life in these species (see Dwyer et al., 1995). Administration of growth hormone (GH) to pregnant sows results in increased secondary muscle fibre number in the piglets at birth, but only in early (10–24 days) gestation when muscle cell proliferation is maximal (Rehfeldt et al., 1993). Administration of GH in mid-gestation results in delayed muscle maturation, whilst administration in late-gestation results in an increased muscle fibre diameter, with no effect on muscle fibre number (i.e. hypertrophic effects – see later). All of these increases in muscle fibre number are via increases in the ratio of secondary-to-primary fibres, due to increased secondary fibres. The number of primary fibres is thought to be determined by genotype, but the number of secondary fibres can be influenced by environmental factors. The mechanism for these effects of environmental factors on muscle fibre number is not known. However, a likely candidate is the fetal GH–IGF axis, which has been shown to be regulated by nutrition in exactly the same way as that of the mother (Bauer et al., 1995). Our recent studies (J.M. Brameld et al., unpublished data) indicate that nutrient restriction of pregnant ewes between 28 and 80 days’ gestation, followed by feeding to control levels results in decreased expression of IGF-I in fetal liver at 80 days, but increased hepatic expression of IGF-II. The effects of nutritional and hormonal manipulations on postnatal growth, and in particular the GH–IGF axis, has recently been reviewed (e.g. Brameld, 1997). We have demonstrated local expression of the IGF-I gene in both skeletal muscle and adipose tissue in the pig, as well as in liver, the major site of production of circulating IGF-I. IGF-I gene expression is affected by diet in a tissue-specific manner (Weller et al., 1994; Brameld et al., 1996b). Evidence from studies in whole animals and cultured hepatocytes suggest that both the protein and energy constituents of a diet have direct effects on hepatic expression of growth-regulatory genes. In pigs, glucose directly increases hepatic expression of the GHR gene and therefore enhances the effects of GH on IGF-I expression, with protein (amino acids) controlling GH-stimulated expression of the IGF-I gene (Brameld et al., 1999a). Interestingly, neither dietary energy nor protein alter IGF-I expression in skeletal muscle, despite having effects on muscle GHR expression opposite to those seen in liver (see Brameld, 1997). This is corroborated by our recent studies which show no effect of nutrient restriction on

Control and Manipulation of Hyperplasia and Hypertrophy

243

IGF-I expression in fetal sheep skeletal muscle, although expression was higher at 80 days’ than at 140 days’ gestation (J.M. Brameld et al., unpublished data). We hypothesized that a possible mechanism for nutrient restriction decreasing muscle fibre number may be via: (i) reduced fetal hepatic IGF-I expression, and therefore reduced mitogenic activity in the blood; and (ii) the reduced proliferation may induce an early peak of IGF-II expression in skeletal muscle, resulting in an early induction of differentiation and therefore reduced muscle fibre number. Indeed, our recent studies of nutrientrestricted ewes indicate an earlier peak of IGF-II expression in skeletal muscle of nutrient-restricted fetal sheep, although only two time points were investigated (J.M. Brameld et al., unpublished data). In some species, including man (Draeger et al., 1987), sheep (Wilson et al., 1992) and pigs (Mascarello et al., 1992), a third generation of muscle fibres have been identified. These tertiary fibres occur during late fetal and very early neonatal development, after the completion of primary and secondary fibre production. Their origin is unclear, but it has been suggested that they are derived from satellite cells (Mascarello et al., 1992) and their development may be affected by nutritional supply (see Dauncey and Gilmour, 1996), since they are present in maternal-fed piglets, but not in piglets fed sow-milk substitute. Whether this is due to differences in the composition of the milk or to the amount of food eaten is still unknown, since maternal-fed animals are likely to have ingested more than the sow-milk substitute-fed animals, which were only fed at 6-hourly intervals (M.J. Dauncey, Cambridge, 1999, personal communication). This suggests that nutrition in late pregnancy and very early neonatal life is important to maximize muscle fibre formation.

Control and manipulation of hypertrophy In postnatal life, increases in muscle size are due to hypertrophy, and not hyperplasia, as muscle fibre number does not increase significantly after birth. Muscle DNA content continues to increase throughout the growing phase due to satellite cell proliferation, differentiation and fusion with existing muscle fibres. This is accompanied by increased protein deposition within the cells. The cessation of DNA accretion occurs abruptly as the animal approaches its mature size and precedes the decline in protein accretion. Muscle protein accretion is the net balance between the relative rates of muscle protein synthesis (i.e. gene transcription and translation into protein) and degradation (proteolysis), and changes in either can result in increased muscle mass. Some of the factors shown to influence cell proliferation and differentiation in cultured cells (Table 14.1) can also influence protein metabolism within muscle cells (Table 14.2). In vitro studies have shown that EGF, insulin and the IGFs all increase protein synthesis and decrease protein breakdown in cultured muscle cells, whereas GH has no effect and dexamethasone increases protein breakdown (Harper and Buttery, 1992).

The use of exogenous hormones Increasing muscle protein accretion, via changes in either protein synthesis or protein degradation (or a combination of the two), forms the basis of several growth promoters

244

Table 14.2. Effects of various growth factors and hormones on differentiated muscle cell protein synthesis and breakdown (see Brameld et al.,1998). Factor Protein synthesis Protein breakdown

Insulin

IGF

GH

FGF

EGF/ TGF-␣

TGF-␤

PDGF

Dex

T3

Test

␤-agonist

TNF-␣

PGE2

PGF2␣

≠ Ø

≠ Ø

Æ Æ

Æ ND

≠ Ø

Ø≠ Ø

≠Æ ND

≠Ø ≠

Æ Æ

Æ Æ

≠Æ Ø

ND ≠

Æ ≠

≠ Æ

IGF, insulin-like growth factors -I and -II; GH, growth hormone; FGF, fibroblast growth factor; EGF/TGF-␣, epidermal growth factor/transforming growth factor-␣ (same receptor); TGF-␤, transforming growth factor-␤; PDGF, platelet-derived growth factor; Dex, Dexamethasone; T3, tri-iodothyronine; Test, testosterone; ␤-agonist, clenbuterol, ractopamine or isoproterenol; TNF-␣, tumour necrosis factor-␣; PGE2, prostaglandin E2; PGF2␣, prostaglandin F2␣; ≠, stimulates; Ø, inhibits; Æ, no effect; ND, no data found.

P.J. Buttery et al.

Control and Manipulation of Hyperplasia and Hypertrophy

245

which have been used to increase muscle mass in farm livestock. The anabolic growth promotors, although no longer available for commercial use in the UK, have been shown to effectively increase muscle mass through hypertrophy. The mode of action of these agents differs, depending on the compound. Longissimus DNA concentration is higher in bulls than in steers (Morgan et al., 1993) and in animals treated with testosterone (Grigsby et al., 1976), indicating an effect of testosterone on satellite cell proliferation. It has been suggested that trenbolone acetate, a testosterone analogue, increases skeletal muscle hypertrophy by enhancing the sensitivity of satellite cells to FGF and IGF-I (Thompson et al., 1989). Testosterone increases both protein synthesis and degradation rates, with a greater effect on the former (Martinez et al., 1984), while trenbolone acetate increases muscle mass predominantly by reducing protein degradation, with a lesser effect on protein synthesis (Vernon and Buttery, 1976). The mode of action of oestradiol is believed to involve enhancement of endogenous GH secretion (Gopinath and Kitts, 1984; Breier et al., 1988), although measurable increases in protein synthetic rate are not always detected in treated animals (Dawson et al., 1991). The response to anabolic agents, however, is clearly dependent on nutritional status (Gill et al., 1987; Bauman et al., 1994). Treating animals with exogenous GH increases the fractional rates of protein synthesis and degradation in skeletal muscle, with the increase in synthesis exceeding the increase in degradation. This results in protein accretion (Pell and Bates, 1987; Eisemann et al., 1989) and therefore hypertrophy. Total muscle RNA concentration is increased in GH-treated muscles suggesting increased protein synthetic capacity rather than increased protein synthesis efficiency (Pell and Bates, 1987). The mechanism for these effects is presumably via endocrine or autocrine/paracrine GH-stimulated IGF-I, since GH has no direct effect on protein metabolism in cultured muscle cells (see above). Although growth hormone administration to pigs increases fibre diameter (Beerman et al., 1990; Rehfeldt and Ender, 1993), its effect on meat quality is variable. Shear force values for samples taken at slaughter were similar for control- and GHtreated pigs, but were higher in GH-treated samples from carcasses hung for 5 days in a cool room (Solomon et al., 1994). This suggests that the post-mortem biochemical processes may be affected by GH, as has been demonstrated in animals treated with ␤agonist (see later). The ambiguity as to whether GH affects meat quality may relate to the magnitude of the effect on proteolysis, since the rate of protein degradation before death affects the rate of protein degradation post mortem, and therefore the tenderization process. Treatment of growing animals with other exogenous agents, such as ␤-adrenergic agonists, also results in increased muscle mass via muscle hypertrophy, with no proliferation of satellite cells. Indeed, muscle DNA concentration (␮g g⫺1 protein) is sometimes less in treated muscles than in control muscles (Kim et al., 1987). The predominant mechanism of action of these agents is believed to be a reduction in protein degradation, although protein synthesis has also been shown to be stimulated in some studies (Dawson et al., 1991). This is supported by the observation that translational efficiency (i.e. the amount of protein synthesized per unit RNA) is increased by ␤-agonists (Maltin et al., 1992). Studies at Nottingham and in the USA have shown that, in sheep and cattle, ␤-adrenergic agonist administration over several weeks has a remarkable effect on the Ca2+-dependent cysteine proteinase (calpain) system (Higgins et al., 1988; Wang and Beermann, 1988; Parr et al., 1992; Speck et al., 1993). The calpain

246

P.J. Buttery et al.

system comprises at least three components, two of which have proteolytic activity in vitro at micromolar and millimolar concentrations of calcium, namely ␮-calpain and m-calpain respectively, with the third important factor being a specific endogenous inhibitor known as calpastatin. ␤-agonist treatment for several weeks results in significant increases in the level of activity of calpastatin and calpastatin mRNA in LD muscle (Higgins et al., 1988; Wang and Beermann, 1988; Parr et al., 1992; Speck et al., 1993). However, not all muscles respond equally to treatment with ␤-agonists (Dawson et al., 1991), probably related to differences in the content of fibre types of different muscles. The cross-sectional area of type II (fast contracting, mixed glycolytic-oxidative) fibres tends to be increased more consistently than type I (slow contracting, oxidative) fibres (see Yang and McElligott, 1989). There also appears to be a change in fibre type proportions in muscles of treated animals following long-term administration of these agents, with a shift towards fibres with increased anaerobic metabolism (Maltin et al., 1986; Zeman et al., 1988). These changes in fibre diameter and metabolism, along with the reduced protein degradation, are believed to contribute to the reduced tenderness of meat treated with ␤-agonists (Aalhus et al., 1992).

Effects of postnatal nutrition Undernutrition has important influences on postnatal growth, as well as on prenatal growth, if it occurs during critical stages of development. Increasing postnatal growth rate via increased dietary protein intake is associated with increased muscle fibre diameter and decreased intramuscular lipid in porcine LD muscle (Essén-Gustavsson et al., 1994). In general, increasing dietary protein above maintenance levels, when the supply of energy is adequate, results in increased protein synthesis and degradation (see Lobley, 1998). The rate of protein synthesis tends to be greater than the rate of degradation and therefore results in a net protein accretion and increased muscle fibre diameter. This increase in diameter has been associated with an increase in meat toughness (see later), although the increased rate of protein degradation would be expected to decrease meat toughness, since the rate of protein degradation before death affects the rate of post-mortem tenderization (see above). Indeed, it has been suggested that a period of ad libitum feeding prior to slaughter increases pork tenderness (Blanchard et al., 1999). It should be stressed that all the effects of exogenous hormones described above are dependent upon an adequate supply of nutrients (see Bauman et al., 1994).

Genetic effects on hypertrophy In sheep, an autosomal dominant gene (callipyge), which is associated with extreme muscling, has been identified on chromosome 18. The effect of this gene is not consistent among all skeletal muscles. The leg and loin muscles all exhibit hypertrophy (18–42%), but certain shoulder muscles (e.g. supraspinatus and infraspinatus) are unaffected (Koohmaraie et al., 1995). The leg and loin muscle hypertrophy exhibited in sheep expressing the callipyge gene is similar to that seen in the double-muscle syndrome in cattle. However, the double-muscle syndrome in cattle is apparent at birth and often results in dystocia, whereas the sheep condition does not manifest itself until

Control and Manipulation of Hyperplasia and Hypertrophy

247

several weeks after birth. This suggests that the mechanisms that give rise to these conditions in cattle differ from those in sheep. The double-muscle syndrome is characterized by increased muscle fibre number, resulting from faster and more prolonged hyperplasia during prenatal growth and development (see above). The callipyge condition is associated with increased muscle DNA content, suggesting greater satellite cell proliferation and increased RNA content, which points to a greater capacity for protein synthesis (Koohmaraie et al., 1995). This condition is thus due to increased hypertrophy rather than hyperplasia. Protein degradation is significantly reduced in these animals, contributing to the increased muscle mass but also reducing meat tenderness, and is due to extremely high levels of calpastatin (Koohmaraie et al., 1995). Fibre type changes are also apparent, with an increase in both proportion and area of the type II fibres and a reduction (or no change) in the size of the red, type I fibres. Sheep with the callipyge gene therefore have 48–62% greater total muscle fibre areas compared with normal lambs (Koohmaraie et al., 1995). These effects on muscle fibre area and fibre type proportions are similar to those often seen in ␤-agonist-treated muscles (Yang and McElligott, 1989), as are the changes in protein synthesis and degradation and meat tenderness.

Meat quality The final eating quality of meat depends on a number of organoleptic properties, including appearance (comprising colour and fat content), taste, texture and tenderness. Whilst colour and fat content are important in influencing meat purchase, consumer studies indicate that it is the degree to which muscle tenderizes after slaughter that is the most important factor contributing to overall meat quality in cattle, sheep and pigs (Koohmaraie, 1994). After slaughter, the loss of oxygen supply to tissues initiates anaerobic metabolism, resulting in the utilization of primary energy stores. In the case of skeletal muscle, this means that muscle glycogen is depleted, producing an increase in lactic acid in the muscle, thereby reducing muscle pH. The rate of postmortem glycogenolysis can be altered in ruminants by circumstances which produce or mimic stress. For example, adrenaline infusion in cattle reduces the rate of pH decrease post mortem by depleting glycogen stores prior to slaughter, giving rise to dark, firm, dry beef (Geesink et al., 1992), with highly fragmented fibres and poor taste qualities. Similar observations have been made on pork carcasses following long-term stress (Warriss et al., 1989), whilst short-term stress increases the rate of pH fall, giving rise to pale, soft, exudative meat. Stress also affects the biochemical processes involved in the tenderization process (see later). The relationship between muscle fibre structure and meat-eating quality is complex, because a number of factors are involved. Attempts to manipulate growth by selective breeding, experimental diets or exogenous growth promoters may change muscle fibre type, mean fibre diameter and the physiological status of the muscle at the time of slaughter. Separating the effects of these factors on tenderness can be difficult, since, for example, glycolytic (white) fibres tend to have larger diameters than oxidative (red) ones. Some studies suggest that the toughness of uncooked meat, as judged from its shear value, may increase with overall fibre diameter. In a trial involving 120 crossbred steers, shear force was positively correlated with fibre diameter, and negatively

248

P.J. Buttery et al.

correlated with the percentage of oxidative fibres (Seideman et al., 1987). Pigs with more muscle fibres also tend to have less fat (Stickland and Goldspink, 1975) while other indices of quality (e.g. toughness) may have also improved. It would therefore seem that high fibre number, which correlates with smaller fibres, may be a relevant parameter that correlates with meat quality. Studies into the effects of dietary protein on tenderness have yielded variable results (see Soloman et al., 1994). EssénGustavsson et al. (1994) found an increase in the toughness of the LD muscle in barrows and gilts fed a high protein diet. These changes were associated with a decrease in intramuscular lipid and an increase in muscle fibre diameter. Among the most important reactions influencing toughness are those catalysed by certain endogenous proteolytic enzymes that act in a highly selective manner on a small number of key intracellular muscle proteins to initiate the tenderization process. Several candidate enzyme systems have been proposed over the last 10 years, including cathepsins and the multicatalytic protease (proteasome). However, in recent years a consensus has emerged that the most important proteolytic enzymes affecting tenderization belong to the calpain system (see above), with the most important substrates being myofibrillar, Z-line and costamere proteins in muscle fibres (Koohmaraie, 1994; Taylor et al., 1995). The calpain system is highly sensitive to fluctuating levels of calcium ion, pH and temperature, all of which change rapidly in the immediate postmortem period (Suzuki et al., 1995). Unlike other proteolytic systems which may be active during the post-mortem period, the calpains are the only proteolytic enzymes in skeletal muscle which do not degrade the major myofibrillar proteins actin and myosin, both of which remain intact during the tenderization process (Goll et al., 1991, 1992). Further evidence has shown that CaCl2 infusion into the carcass increases the rate of tenderization in beef whilst infusion of ZnCl2, a calpain inhibitor, reduces the rate of tenderization in both beef and lamb (Koohmaraie, 1990; Geesink et al., 1994). Longterm ␤-agonist treatment is known to produce tough meat in cattle and sheep, presumably due to suppression of post-mortem proteolysis by the elevated activity of calpastatin (Kretchmar et al., 1990; Wheeler and Koohmaraie, 1992). The rate of tenderization also varies between the principal meat-producing species, where (in order of greater tenderization) pork > lamb > beef. Significantly, calpastatin levels at slaughter are greatest in beef and lowest in pork, adding further evidence to the involvement of the calpain system in the tenderization process (Koohmaraie et al., 1991). Callipyge lambs, characterized by their enhanced muscle growth and excessively tough meat, have extremely high levels of calpastatin (Koohmaraie et al., 1995). Whilst the meat toughness is not further enhanced by ␤-agonist treatment, it can be reduced by infusion of CaCl2 (Koohmaraie et al., 1996; Clare et al., 1997). Thus, a genetic condition has been shown to alter the calpain system in a manner consistent with its effects on tenderization. We have shown that the same system is involved in the conversion of pig muscle to pork and that pre-slaughter stress can alter the post-mortem responsiveness of the calpain system (Sensky et al., 1996, 1999; Parr et al., 1999a, b).

Acknowledgements The authors gratefully acknowledge the financial support of the BBSRC, Meat and Livestock Commission and the MAFF for much of their research.

Control and Manipulation of Hyperplasia and Hypertrophy

249

References Aalhus, J.L., Schaefer, A.L., Murray, A.C. and Jones, S.D.M. (1992) The effect of ractopamine on myofiber distribution and morphology and their relation to meat quality in swine. Meat Science 31, 397–409. Allen, R.E., Luiten, L.S. and Dodson, M.V. (1985) Effect of insulin and linoleic acid on satellite cell differentiation. Journal of Animal Science 60, 1571–1579. Arnold, H.H., Gerharz, C.D., Gabbert, H.E. and Salminen, A. (1992) Retinoic acid induces myogenin synthesis and myogenic differentiation in the rat rhabdomyosarcoma cell line BA-Han-1C. Journal of Cell Biology 118, 877–887. Bauer, M.K., Breier, B.H., Harding, J.E., Veldhuis, J.D. and Gluckman, P.D. (1995) The fetal somatotropic axis during long term maternal undernutrition in sheep: Evidence for nutritional regulation in utero. Endocrinology 136, 1250–1257. Bauman, D.E., Beerman, D.H., Boyd, R.D., Buttery, P.J., Campbell, R.B., Chalupa, W.V., Etherton, T.D., Klasing, K., Schelling, G.T. and Steele, N.C. (1994) Metabolic Modifiers: Effects on the Nutrient Requirements of Food-Producing Animals. National Academy Press, Washington USA, 81 pp. Beerman, D.H., Fishell, V.K., Roneker, K., Boyd, R.D., Armbruster, G. and Souza, L. (1990) Dose-response relationships between porcine somatotropin, muscle composition, muscle fiber characteristics and pork quality. Journal of Animal Science 68, 2690–2697. Blanchard, P.J., Ellis, M., Warkup, C.C., Hardy, B., Chadwick, J.P. and Deans, G.A. (1999) The influence of rate of lean and fat tissue development on pork eating quality. Animal Science 68, 477–485. Brameld, J.M. (1997) Molecular mechanisms involved in the nutritional and hormonal regulation of growth in pigs. The Proceedings of the Nutrition Society 56, 607–619. Brameld, J.M., Atkinson, J.L., Budd, T.J., Saunders, J.C., Pell, J.M., Salter, A.M., Gilmour, R.S. and Buttery, P.J. (1996a) Expression of insulin-like growth factor-I (IGF-I) and growth hormone receptor (GHR) mRNA in liver, skeletal muscle and adipose tissue of different breeds of pig. Animal Science 62, 555–559. Brameld, J.M., Atkinson, J.L., Saunders, J.C., Pell, J.M., Buttery, P.J. and Gilmour, R.S. (1996b) Effects of Growth Hormone administration and dietary protein intake on Insulin-like growth factor-I and Growth Hormone Receptor mRNA expression in porcine liver, skeletal muscle and adipose tissue. Journal of Animal Science 74, 1832–1841. Brameld, J.M., Buttery, P.J., Dawson, J.M. and Harper, J.M.M. (1998) Nutritional and hormonal control of skeletal muscle cell growth and differentiation. Proceedings of the Nutrition Society 57, 207–217. Brameld, J.M., Gilmour, R.S. and Buttery, P.J. (1999a) Glucose and amino acids interact with hormones to control expression of insulin-like growth factor-I (IGF-I) and growth hormone receptor (GHR) mRNA in cultured pig hepatocytes. Journal of Nutrition 129, 1298–1306. Brameld, J.M., Smail, H., Imram, N., Millard, N. and Buttery, P.J. (1999b) Changes in expression of IGF-I, IGF-II and GH receptor (GHR) mRNA during differentiation of cultured muscle cells derived from adult and fetal sheep. Proceedings of the IXth International Society for Ruminant Physiology, Pretoria, South Africa (in press). Breier, B.H., Gluckman, P.D. and Bass, J.J. (1988) Influence of nutritional status and oestradiol17ß on plasma growth hormone, insulin-like growth factors-I and -II and the response to exogenous growth hormone in young steers. Journal of Endocrinology 118, 243–250. Brunetti, A. and Goldfine, I.D. (1990) Role of myogenin in myoblast differentiation and its regulation by fibroblast growth factor. Journal of Biological Chemistry 265, 5960–5963. Clare, T.L., Jackson, S.P., Miller, M.F., Elliott, C.T. and Ramsey, C.B. (1997) Improving tenderness of normal and callipyge lambs with calcium chloride. Journal of Animal Science 75, 377–385.

250

P.J. Buttery et al. Dauncey, M.J. and Gilmour, R.S. (1996) Regulatory factors in the control of muscle development. Proceedings of the Nutrition Society 55, 543–559. Dawson, J.M., Buttery, P.J., Lammiman, M.J., Soar, J.B., Essex, C.P., Gill, M. and Beever, D.E. (1991) Nutritional and endocrinological manipulation of lean deposition in forage-fed steers. British Journal of Nutrition 66, 171–185. Desai, M. and Hales, C.N. (1997) Role of fetal and infant growth in programming metabolism in later life. Biological Reviews 72, 329–348. Dodson, M.V., Mathison, B.A. and Mathison, B.D. (1990) Effects of medium and substratum on ovine satellite cell attachment, proliferation and differentiation in vitro. Cell Differentiation and Development 29, 59–66. Doumit, M.E. and Merkel, R.A. (1992) Conditions for isolation and culture of porcine myogenic satellite cells. Tissue and Cell 24, 253–262. Downes, M., Mynett-Johnson, L. and Muscat, G.E.O. (1994) The retinoic acid and retinoid X receptors are differentially expressed during myoblast differentiation. Endocrinology 134, 2658–2661. Draeger, A., Weeds, A.G. and Fitzsimons, R.B. (1987) Primary, secondary and tertiary myotubes in developing skeletal muscle: a new approach to the analysis of human myogenesis. Journal of Neurological Science 81, 19–43. Dwyer, C.M. and Stickland, N.C. (1994) Supplementation of a restricted maternal diet with protein or carbohydrate alone prevents a reduction in fetal muscle fibre number in the guinea-pig. British Journal of Nutrition 72, 173–180. Dwyer, C.M., Madgwick, A.J.A., Crook, A.R. and Stickland, N.C. (1992) The effect of maternal undernutrition on the growth and development of the guinea-pig placenta. Journal of Developmental Physiology 18, 295–302. Dwyer, C.M., Fletcher, J.M. and Stickland, N.C. (1993) Muscle cellularity and postnatal growth in the pig. Journal of Animal Science 71, 3339–3343. Dwyer, C.M., Stickland, N.C. and Fletcher, J.M. (1994) The influence of maternal nutrition on muscle fiber number development in the porcine fetus and on subsequent postnatal growth. Journal of Animal Science 72, 911–917. Dwyer, C.M., Madgwick, A.J.A., Ward, S.S. and Stickland, N.C. (1995) Effect of maternal undernutrition in early gestation on the development of fetal myofibres in the guinea-pig. Reproduction, Fertility and Development 7, 1285–1292. Eisemann, J.H., Hammond, A.C. and Rumsey, T.S. (1989) Tissue protein synthesis and nucleic acid concentrations in steers treated with somatotropin. British Journal of Nutrition 62, 657–671. Essén-Gustavsson, B., Karlsson, K., Lundström, K. and Enfält, A-C. (1994) Intramuscular fat and muscle fiber lipid contents in halothane-gene-free pigs fed high or low protein diets and its relation to meat quality. Meat Science 38, 269–277. Ezekwe, M.O. and Martin, R.J. (1975) Cellular characteristics of skeletal muscle in selected strains of pigs and mice and the unselected controls. Growth 39, 95–106. Gal-Levi, R., Leshem, Y., Aoki, S., Nakamura, T. and Halevy, O. (1998) Hepatocyte growth factor plays a dual role in regulating skeletal muscle satellite cell proliferation and differentiation. Biochimica et Biophysica Acta 1402, 39–51. Geesink, G.H., Ouali, A., Smulders, F.J.M., Talmant, A., Tassy, C., Guignot, F. and van Laack, H.L.J.M. (1992) The role of ultimate pH in proteolysis and calpain calpastatin activity in bovine muscle. Biochimie 74, 283–289. Geesink, G.H., Smulders, F.J.M. and Van Laack, R.L.J.M. (1994) The effects of calciumchlorides, sodium-chlorides and zinc-chlorides treatment on the quality of beef. Sciences des Aliments 14, 485–502. Gerrard, D.E. and Grant, A.L. (1994) Insulin-like growth factor-II expression in developing skeletal muscle of double muscled and normal cattle. Domestic Animal Endocrinology 11, 339–347.

Control and Manipulation of Hyperplasia and Hypertrophy

251

Gerrard, D.E. and Judge, M.D. (1993) Induction of myoblast proliferation in L6 myoblast cultures by fetal serum of double muscled and normal cattle. Journal of Animal Science 71, 1464–1470. Gill, M., Beever, D.E., Buttery, P.J., England, P., Gibb, M.J. and Baker, R.D. (1987) The effect of oestradiol-17␤ implantation on the response in voluntary intake, live-weight gain and body composition, to fishmeal supplementation of silage offered to growing calves. Journal of Agricultural Science, Cambridge 108, 9–16. Goll, D.E., Dayton, W.R., Singh, I. and Robson, R.M. (1991) Studies of the alpha-actinin actin interaction in the z-disk by using calpain. Journal of Biological Chemistry 266, 8501–8510. Goll, D.E., Thompson, V.F., Taylor, R.G. and Christiansen, J.A. (1992) Role of the calpain system in muscle growth. Biochimie 74, 225–237. Gopinath, R. and Kitts, W.D. (1984) Growth hormone secretion and clearance rate in growing beef steers implanted with oestrogenic anabolic compounds. Growth 48, 499–514. Grigsby, J.S., Bergen, W.G. and Merkel, R.A. (1976) The effect of testosterone on skeletal muscle development and protein synthesis in rabbits. Growth 40, 303–316. Handel, S.E. and Stickland, N.C. (1987) Muscle cellularity and birth weight. Animal Production 44, 311–317. Harper, J.M.M. and Buttery, P.J. (1992) Muscle cell growth. In: Boorman, K.N., Buttery, P.J. and Lindsay, D.B. (eds) The Control of Fat and Lean Deposition. Butterworths, London, pp. 27–58. Hegarty, P.V.J. and Allen, C.E. (1978) Effect of pre-natal runting on the post-natal development of skeletal muscle in swine and rats. Journal of Animal Science 46, 1634–1640. Higgins, J.A., Lasslett, Y.V., Bardsley, R.G. and Buttery, P.J. (1988) The relation between dietary restriction or clenbuterol (a selective ␤-2 agonist) treatment on muscle growth and calpain proteinase (ec3.4.22.17) and calpastatin activities in lambs. British Journal of Nutrition 60, 645–652. Hong, D.H. and Forsberg, N.E. (1994) Effects of serum and insulin-like growth factor I on protein degradation and protease gene expression in rat L8 myotubes. Journal of Animal Science 72, 2279–2288. Kim, Y.S., Lee, Y.B. and Dalymple, R.H. (1987) Effect of the repartitioning agent, cimaterol on growth, carcass and skeletal muscle characteristics in lambs. Journal of Animal Science 65, 1392–1399. Koohmaraie, M. (1990) Inhibition of postmortem tenderization in ovine carcasses through infusion of zinc. Journal of Animal Science 68, 1476–1483. Koohmaraie, M. (1994) Muscle proteinases and meat aging. Meat Science 36, 93–104. Koohmaraie, M., Whipple, G., Kretchmar, D.H., Crouse, J.D. and Mersmann, H.J. (1991) Postmortem proteolysis in longissimus muscle from beef, lamb and pork carcasses. Journal of Animal Science 69, 617–624. Koohmaraie, M., Shackelford, S.D., Wheeler, T.L., Longergan, S.M. and Doumit, M.E. (1995) A muscle hypertrophy condition in lamb (callipyge): characterization of effects on muscle growth and meat quality traits. Journal of Animal Science 73, 3596–3607. Koohmaraie, M., Shackelford, S.D. and Wheeler, T.L. (1996) Effects of a beta-adrenergic agonist (L-644,969) and male sex condition on muscle growth and meat quality of callipyge lambs. Journal of Animal Science 74, 70–79. Kretchmar, D.H., Hathaway, M.R., Epley, R.J. and Dayton, W.R. (1990) Alterations in postmortem degradation of myofibrillar proteins in muscle of lambs fed a ␤-adrenergic agonist. Journal of Animal Science 68, 1760–1772. Kwak, K.B., Chung, S.S., Kim, O.M., Kang, M.S., Ha, D.B. and Chung, C.H. (1993) Increase in the level of m-calpain correlates with the elevated cleavage of filamin during myogenic differentiation of embryonic muscle cells. Biochimica et Biophysica Acta 1175, 243–249. Lobley, G.E. (1998) Nutritional and hormonal control of muscle and peripheral tissue metabolism in farm species. Livestock Production Science 56, 91–114.

252

P.J. Buttery et al. McCoard, S.A., Peterson, S.W., McNabb, W.C., Harris, P.M. and McCutcheon, S.N. (1997) Maternal constraint influences muscle fibre development in fetal lambs. Reproduction, Fertility and Development 9, 675–681. McPherron, A.C., Lawler, A.M. and Lee, S.J. (1997) Regulation of skeletal muscle mass in mice by a new TGF-␤ superfamily member. Nature 387, 83–90. Maltin, C.A., Delday, M.I. and Reeds, P.J. (1986) The effect of a growth promoting drug, clenbuterol, on fibre frequency and area in hind limb muscles from young rats. Bioscience Reports 6, 293–299. Maltin, C.A., Hay, S.M., McMillan, D.N. and Delday, M.I. (1992) Tissue specific responses to clenbuterol: temporal changes in protein metabolism of striated muscle and visceral tissues from rats. Growth Regulation 2, 161–166. Martinez, J.A., Buttery, P.J. and Pearson, J.T. (1984) The mode of action of anabolic agents, the effects of testosterone on growth rates and muscle protein metabolism in female rats. British Journal of Nutrition 54, 515–521. Mascarello, F., Stecchini, M.L., Rowlerson, A. and Ballocchi, E. (1992) Tertiary myotubes in postnatal growing pig muscle detected by their myosin isoform composition. Journal of Animal Science 70, 1806–1813. Maxfield, E.K., Sinclair, K.D., Broadbent, P.J., McEvoy, T.G., Robinson, J.J. and Maltin, C.A. (1998) Short-term culture of ovine embryos modifies fetal myogenesis. American Journal of Physiology 274, E1121-E1123. Miller, L.R., Garwood, V.A. and Judge, M.D. (1975) Factors affecting porcine muscle fiber type, diameter and number. Journal of Animal Science 41, 66–77. Moore, J.W., Dionne, C., Jaye, M. and Swain, J.L. (1991) The mRNAs encoding acidic FGF, basic FGF and FGF receptor are coordinately downregulated during myogenic differentiation. Development 111, 741–748. Morgan, J.B., Wheeler, T.L., Koohmaraie, M., Crouse, J.D. and Savell, J.W. (1993) Effect of castration on myofibrillar protein turnover, endogenous proteinase activities and muscle growth in bovine skeletal muscle. Journal of Animal Science 71, 408–414. Morris, G.E., Piper, M. and Cole, R. (1976) Differential effects of calcium ion concentration on cell fusion, cell division and creatine kinase activity in muscle cell cultures. Experimental Cell Research 99, 106–114. Parr, T., Bardsley, R.G., Gilmour, R.S. and Buttery, P.J. (1992) Changes in calpain and calpastatin messenger-RNA induced by ␤-adrenergic stimulation of bovine skeletal-muscle. European Journal of Biochemistry 208, 333–339. Parr, T., Sensky, P.L., Scothern, G., Bardsley, R.G., Buttery, P.J., Wood, J.D. and Warkup, C. (1999a) Immunochemical study of the calpain system in porcine longissimus muscle with high and low shear force values. Journal of Animal Science 77(suppl. 1), 164. Parr, T., Sensky, P.L., Warkup, C., Bardsley, R.G. and Buttery, P.J. (1999b) Changes in porcine skeletal muscle calpastatin expression induced by a single dose of a ␤2-agonist. Journal of Animal Science 77(suppl. 1), 164–165. Pell, J.M. and Bates, P.C. (1987) Collagen and non-collagen protein turnover in skeletal muscle of growth hormone-treated lambs. Journal of Endocrinology 115, R1-R4. Petrie, L., Buskin, J.N. and Chesters, J.K. (1996) Zinc and the initiation of myoblast differentiation. Journal of Nutritional Biochemistry 7, 670–676. Powell, S.E. and Aberle, E.D. (1981) Skeletal-muscle and adipose-tissue cellularity in runt and normal birth-weight swine. Journal of Animal Science 52, 748–756. Quinn, L.S., Ong, L.D. and Roeder, R.A. (1990) Paracrine control of myoblast proliferation and differentiation by fibroblasts. Developmental Biology 140, 8–19. Rehfeldt, C. and Ender, K. (1993) Skeletal-muscle cellularity and histochemistry in response to porcine somatotropin in finishing pigs. Meat Science 34, 107–118. Rehfeldt, C., Fiedler, I., Weikard, R., Kanitz, E. and Ender, K. (1993) It is possible to increase skeletal-muscle fiber number in utero. Bioscience Reports 13, 213–220.

Control and Manipulation of Hyperplasia and Hypertrophy

253

Robinson, J.J., Sinclair, K.D. and McEvoy, T.G. (1999) Nutritional effects on foetal growth. Animal Science 68, 315–331. Rosenthal, S.M., Brown, E.J., Brunetti, A. and Goldfine, I.D. (1991) Fibroblast growth factor inhibits insulin-like growth factor-II (IGF-II) gene expression and increases IGF-I receptor abundance in BC3H-1 muscle cells. Molecular Endocrinology 5, 678–684. Seideman, S.C., Koohmaraie, M. and Crouse, J.D. (1987) Factors associated with tenderness in young beef. Meat Science 20, 281–291. Sensky, P.L., Parr, T., Bardsley, R.G. and Buttery, P.J. (1996) The relationship between plasma epinephrine concentration and the activity of the calpain enzyme-system in porcine longissimus muscle. Journal of Animal Science 74, 380–387. Sensky, P.L., Parr, T., Lockley, A.K., Bardsley, R.G., Buttery, P.J., Wood, J.D. and Warkup, C. (1999) Altered calpain levels in longissimus muscle from normal pigs and heterzygotes with the ryanodine receptor mutation. Journal of Animal Science 77, 2956–2964. Sinclair, K.D., McEvoy, T.G., Carolan, C., Maxfield, E.K., Maltin, C.A., Young, L.E., Wilmut, I., Robinson, J.J. and Broadbent, P.J. (1998) Conceptus growth and development following in vitro culture of ovine embryos in media supplemented with bovine serum. Theriogenology 49, 218. Solomon, M.B., Caperna, T.J., Mroz, R.J. and Steele, N.C. (1994) Influence of dietary protein and recombinant porcine somatotropin administration in young pigs, III. Muscle fiber morphology and shear force. Journal of Animal Science 72, 615–621. Speck, P.A., Collingwood, K.M., Bardsley, R.G., Tucker, G.A., Gilmour, R.S. and Buttery, P.J. (1993) Transient changes in growth and in calpain and calpastatin expression in ovine skeletal-muscle after short-term dietary inclusion of cimaterol. Biochimie 75, 917–923. Stickland, N.C. and Goldspink, G. (1975) A note on porcine skeletal muscle parameters and their possible use in early progeny testing. Animal Production 21, 93–96. Summers, P.J. and Medrano, J.F. (1997) Delayed myogenesis associated with muscle fiber hyperplasia in High-Growth mice. Proceedings of the Society for Experimental Biology and Medicine 214, 380–385. Suzuki, K., Sorimachi, H., Yoshizawa, T., Kinbara, K. and Ishiura, S. (1995) Calpain, novel family members, activation, and physiological function. Biological Chemistry Hoppe-Seyler 376, 523–529 Taylor, R.G., Geesink, G.H., Thompson, V.F., Koohmaraie, M. and Goll, D.E. (1995) Is z-disk degradation responsible for postmortem tenderization. Journal of Animal Science 73, 1351–1367. Thompson, S.H., Boxhorn, L.K., Wong, W. and Allen, R.E. (1989) Trenbolone alters the responsiveness of skeletal muscle satellite cells to fibroblast growth factor and insulin-like growth factor I. Endocrinology 124, 2110–2117. Ueda, Y., Wang, M.-C., Ou, B.R., Huang, J., Elce, J., Tanaka, K., Ichihara, A. and Forsberg, N.E. (1998) Evidence for the participation of the proteasome and calpain in early phases of muscle cell differentiation. International Journal of Biochemistry and Cell Biology 30, 679–694. Vernon, B.G. and Buttery, P.J. (1976). Protein turnover in rats treated with trenbolone acetate. British Journal of Nutrition 36, 575–579. Wang, S.Y. and Beermann, D.H. (1988) Reduced calcium-dependent proteinase activity in cimaterol-induced muscle hypertrophy in lambs. Journal of Animal Science 66, 2245–2550. Warriss, P.D., Bevis, E.A. and Ekins, P.J. (1989) The relationships between glycogen stores and muscle ultimate pH in commercially slaughtered pigs. British Veterinary Journal 145, 378–383. Weller, P.A., Dauncey, M.J., Bates, P.C., Brameld, J.M., Buttery, P.J. and Gilmour, R.S. (1994) Regulation of porcine insulin-like growth factor-I and growth hormone receptor mRNA expression by energy status. American Journal of Physiology 266, E776–E785. Wheeler, T.L. and Koohmaraie, M. (1992) Effects of the ß-adrenergic agonist L644,969 on muscle protein-turnover, endogenous proteinase activities, and meat tenderness in steers. Journal of Animal Science 70, 3035–3043.

254

P.J. Buttery et al. Wilson, S.J., Ross, J.J. and Harris, A.J. (1988) A critical period for formation of secondary myotubes defined by prenatal undernourishment in rats. Development 102, 815–821. Wilson, S.J., McEwan, J.C., Sheard, P.W. and Harris, A.J. (1992) Early stages of myogenesis in a large mammal: formation of successive generations of myotubes in sheep tibialis cranialis muscle. Journal of Muscle Research and Cell Motility 13, 534–550. Yang, Y.T. and McElligott, M.A. (1989) Multiple actions of ␤-adrenergic agonists on skeletal muscle and adipose tissue. Biochemical Journal 261, 1–10. Zeman, R.J., Ludemann, R., Easton, T.G. and Etlinger, J.D. (1988) Slow to fast alterations in skeletal muscle fibers caused by clenbuterol, a ␤2-receptor agonist. American Journal of Physiology 254, E726–E732.

15

Regulation of Protein Synthesis for Wool Growth

N.R. ADAMS, S. LIU AND D.G. MASTERS CSIRO Division of Animal Production and CRC for Premium Quality Wool, Wembley, Western Australia

Introduction Despite numerous studies on wool growth, there is not yet an integrated biological understanding of how wool growth rate is controlled. Genetic and nutritional effects on wool growth have been studied intensively, but the biological mechanisms underlying the substantial interactions between these factors remain ill-defined (Jackson and Roberts, 1970; Woolaston, 1987). This chapter aims to develop an integrative framework that will facilitate collaboration between nutritionists, physiologists, and cell and molecular biologists to develop practical systems to improve wool growth. Such a framework is needed, for example, to assess the likely value of individual genes for wool growth, as new information becomes available. Subsequent work is planned to use the framework to develop more quantitative applied models relating wool growth rate to staple strength, meat production and reproduction. We have approached this task by considering follicles as scattered units of a single organ that produces wool, and treating the organ as a whole (much as milk production is normally considered as a function of the mammary gland, rather than of individual mammary alveoli). This contrasts with the more traditional approach of considering wool growth as an aggregation of the characteristics of individual follicles, and describing wool growth by detailed analysis of follicle biology. Although our approach glosses over much of the detailed information about function within the follicle, it enables the integration of follicle function with the overall protein synthetic activity in the skin, and in turn, in the body. This approach is possible because wool growth represents a relatively constant proportion of protein synthesis in skin (Table 15.1). Figure 15.1 presents the framework developed in this chapter. Wool growth rate is considered as a function of the mass of follicular tissue, the rate at which this tissue synthesizes protein, and the proportion of that protein extruded as wool. In this schema, genotype may affect the capacity for wool growth at a number of sites, including skin mass, the fraction of skin made up of follicular tissue, the capacity of follicles to use amino acids, follicle efficiency, uptake of amino acids by other tissues, © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

255

256

N.R. Adams et al. Table 15.1. The proportion of skin protein synthesis deposited into wool (skin efficiency) in sheep with different fractional synthesis rates (FSR) of protein in the skin. Dietary level (M)

FSR (% day⫺1)

Romney ewes

1.4 0.6

15.3 5.6

0.15 0.14

Harris et al. (1994b)

Merino lambs

1 0.6

17.3 15.3

0.18 0.20

Liu et al. (1998)

Merino wethers, Fleece+ Fleece⫺

1 1

18.2 14.4

0.19 0.16

D.G. Masters et al. (unpublished)

Merino ewes, dry Pregnancy Lactation

1 1 1

11.7 10.1 11.9

0.16–0.21a 0.10a 0.16a

Sheep

Skin efficiency for wool Source

S. Liu et al. (unpublished)

a

Ewes fed to maintain empty body weight. Skin efficiency was calculated from wool growth (Masters and Mata, 1996) and estimated skin protein content.

Wool growth Efficiency (wool/skin) (G, H) Follicle mass

Skin protein synthesis rate

Sensitivity and responsiveness to nutrients (G) Fraction of skin as follicles (G, H)

Skin mass (G)

Supply of amino acids to skin Use of amino acids by other tissues (G, H, N)

Previous nutrition (N)

Supply of amino acids (N)

Fig. 15.1. Diagrammatic model of the control of wool growth, indicating points at which genetic (G), hormonal (H) and nutritional (N) control may be exerted.

Regulation of Protein Synthesis for Wool Growth

257

responsiveness of the body to nutrient supply, and feed intake. Nutrition may affect wool growth in one of two ways. Firstly, it may increase total mass of follicular tissue. As with other body tissues (e.g. skin or muscle mass), such a change will depend primarily on the level of feed intake. Secondly, the rate of protein synthesis in follicles depends primarily on the supply of amino acids. Hormonal and physiological state may modulate and mediate these effects at several sites.

Models of wool growth The growth of wool can be conceptualized in three main ways. Efforts to integrate wool growth rate with wool fibre qualities have focused on the characteristics of individual wool follicles. Prediction of wool growth under practical husbandry conditions has utilized nutrient supply models. In this review, protein synthesis rate is used to provide a basis to integrate both nutritional and genetic mechanisms affecting wool growth.

Follicle models There has been considerable work relating cell dynamics in the wool follicle to wool growth. The pioneering model in this area used the supply of amino acids to the follicle to drive the rate of mitosis in the follicle bulb, which had a genetic limit to its rate of response (Black and Reis, 1979). This concept has been developed further (e.g. Scobie and Woods, 1992; Bowman et al., 1993). Subsequent work showed that enhanced nutrient supply not only increased mitotic rate (or decreased cell turnover time), but also increased the number of cells in the bulb germinal epithelium, and increased the volume of fibre cortical cells (Table 15.2). The relative importance of these three components of wool growth varied, presumably depending on the diet and duration or rate of feeding, but this cannot be determined from these studies. Although the role of cell apoptosis in determining the availability of cells to the fibre is unknown, the total number of cells generated (i.e. the product of bulb volume and mitotic rate) correlated well with changes in wool fibre production, while cortical cell size played a less important role (Table 15.2). Genetic differences in fleece weight appear to be independent of cell turnover time (Williams and Winston, 1987; Hynd, 1989), but are also reflected in the amount of germinal tissue in follicle bulbs, which in turn is a product of bulb volume and follicle density. Seasonal changes in wool growth in Romney rams were also independent of changes in cell turnover time (Holle et al., 1994). Thus, despite the conclusion of Wilson and Short (1979), cell turnover time alone does not provide a reliable means for modelling wool growth. Total cell production rate by the bulb correlates closely with wool growth rate, but this does not prove causation. As with many other tissues, the rate of cell production by the follicle is inversely related to the rate of cell differentiation, so that cells labelled with [3H]thymidine take longer to differentiate and mature as the cell production rate increases (Chapman et al., 1980). When wool production rate was doubled as a result of increased feed intake, the time cells took to pass through the keratinization zone of

258

N.R. Adams et al.

Table 15.2. Percentage changes in fibre production and follicle characteristics associated with increased feeding in Merino sheep.

Duration of study (days)

Increase in cell turnover rate

Increase in bulb volume

Increase in cortical cell volume

250

28

47

12

19

164

365

20

22

33

130

106

0

3

0

133 186

56 49

6 25

30 94

5 76

Increase in wool growth

Reference Short et al. (1968) Wilson and Short (1979) Williams and Winston (1987) Hynd (1989) Hynd (1994b)

the follicle increased from 19 to 25 h, whilst the time taken for incorporation of [35S]cystine into insoluble keratin doubled from 7 to 14.5 h (Chapman et al., 1980). It is therefore equally sustainable to hypothesize that wool growth rate is limited by the capacity for cell maturation in the developing fibre shaft. Robust and complex control mechanisms ensure close coordination between cell production, cell differentiation and cell maturation in the wool follicle. This makes it impossible to nominate any one of these functions as the ultimate controller of wool growth, because they all correlate closely with each other and also with wool growth. Nevertheless, the simplest hypothesis is that nutritional control of wool growth is exerted by affecting the capacity of the follicle to support protein synthesis for cell maturation, rather than by direct control of cell production in the bulb.

Nutritional models The earliest nutritional models were simply equations derived to show a linear relationship between feed intake and wool growth. These were later adjusted to linear relationships between wool growth and digestible dry matter intake, to take account of variation in feed quality between experiments (Allden, 1979). However, wool growth is even more closely related to the amount of protein absorbed from the small intestine (Black et al., 1973). The amount of protein arriving in the small intestine is a function of the fermentable energy available for microbial protein synthesis, the protein content of the diet and the extent to which that protein is degraded in the rumen. The most successful model to date (GrazFeed; Freer et al., 1997) estimates ‘Digestible Protein Leaving the Stomach’ from a range of feed and intake variables, including those described above. This model achieves accurate prediction in many circumstances and has been applied widely in practical animal husbandry to calculate appropriate levels of pasture and supplement for sheep. However, differences between genotypes are dealt with by

Regulation of Protein Synthesis for Wool Growth

259

classifying sheep broadly by age, sex and breed type, and therefore the model does not directly account for interactions between genotype and nutrient supply. In addition, GrazFeed predicts wool growth as a constant proportion of protein absorbed, regardless of the composition of that protein. However, the amino acid composition of the absorbed protein can have a profound effect on the efficiency of conversion of protein to wool. For example, the low level of sulphur amino acids in microbial protein usually limits wool growth, and provision of additional methionine may increase efficiency of conversion of absorbed protein to wool by 30% (Mata and Masters, 1999). It is likely that nutritional models will develop further to include absorption and use of specific amino acids for wool. The rate of blood flow to the skin regulates the nutrient supply to the follicle, but it is almost certainly not the primary control mechanism (Hales and Fawcett, 1993). This does not mean it is not important. Changes in wool growth rate are usually accompanied by changes in blood flow to the skin, and Harris et al. (1994b) suggested that changes in blood flow were as important in determining the rate of protein synthesis in the skin as they are in mammary gland.

Protein synthesis Most nutritional models focus on protein deposition, which is the difference between protein synthesis rate and protein degradation rate. Thus, they model mass balance, rather than rate constants. However, wool protein is excreted by the wool follicle so it does not undergo ‘turnover’. Wool growth therefore depends only on protein synthesis rate. Models of protein turnover rates are in their infancy (Sainz and Wolff, 1990; Knap and Schrama, 1996), but offer a basis for specialized wool models of the future. If we consider wool growth as part of the function of skin as whole, a simple model to relate wool growth to protein synthesis in skin can be described as follows: Wool growth = Protein content (Pc) ⫻ Synthesis rate (Ps) ⫻ Skin efficiency (r) = [Pc0 + Pc0 ⫻ (Ps ⫺ Pd)] ⫻ Ps ⫻ r Pc0 may be defined as Pc at a maintenance state, Pd is protein degradation rate in skin and r is the proportion of skin protein synthesis committed into wool. We can replace Pc with [Pc0 + Pc0 ⫻ (Ps – Pd)] to allow for the change in skin protein mass, which is determined by both synthesis and degradation rates. Based on this model, wool growth will reflect three main components: skin (protein) mass and its change, synthesis rate and skin efficiency. The following discussion explores each of these components in greater detail, as a basis for developing such a model.

Control of skin and follicular mass The mass of follicular tissue depends on both the fraction of the skin made up by follicles, and on the mass of skin. During normal growth, changes in skin mass generally follow changes in the mass of the whole body. For example, as Suffolk-Finn Dorset lambs grew from 25 kg to 40 kg and then 55 kg, the proportion of nitrogen in skin to the total nitrogen in the whole body was almost constant at 0.091, 0.11 and 0.106

260

N.R. Adams et al.

(MacRae et al., 1993, recalculated to exclude wool N). However, skin mass can be altered differentially by the level of nutrition.

Nutrition As shown in Table 15.2, changes in the amount of germinal epithelium in skin contribute to the effect of nutrition on wool growth. With increased levels of feed intake, both the weight of skin per unit area and the amount of non-collagenous protein per unit mass increase (Williams and Morley, 1994). However, there are no reliable data on the effects of nutrition on the proportion of skin mass made up by follicles. While skin and muscle mass generally move in the same direction when feed intake changes, the overall rates and amplitudes of change can be quite different. For example, Murray and Slezacek (1994) reported a 35% reduction in skin weight, compared with a 13% reduction in muscle and a 53% reduction in alimentary tract, associated with an 18% body weight loss in young sheep. Similarly, when Merino hoggets were fed at 0.44 M, 1.15 M or 1.8 M, the carcass weight increased by 29% from the lowest to highest intake, while the skin weight increased by 65% (N.R. Adams, unpublished observations). The protein concentration in skin did not change with intake, so the weight change indicated similar changes in protein mass in the skin. Although changes in skin mass are relatively large, the rate of change is slower than changes in fractional protein synthesis rate (Liu et al., 1998) or in the mitotic rate of bulb cells (A.C. Schlink, unpublished observations). The slow rate of change in mass would contribute to the delayed response of wool growth to changes in nutrition summarized by Nagorka (1977). In addition to slow changes in follicle mass, the substantial delays in wool growth response to re-feeding in sheep that have been subjected to restricted nutrition (Butler-Hogg, 1984) are likely to be affected also by diversion of amino acids to tissues with a greater demand (see p. 264).

Genetics Genetic factors affect the mass of follicles mainly through effects on the proportion of skin mass made up of follicles. Jackson et al. (1975) calculated that 83% of genetic difference between sheep in fleece weight was correlated with four descriptors of follicle morphology: follicle length, follicle curvature, the number of follicles per unit area of skin, and the ratio of primary to secondary follicles. Subsequent work supports these observations. Hocking Edwards and Hynd (1992) found that the relative volume of germinative tissue in skin of sheep is correlated with their genetic capacity for wool growth, and Williams and Winston (1987) found that the major difference between sheep that differed genetically for fleece weight was the length of follicles, rather than their diameter. Genetic effects on wool production through total skin mass are more limited. Low and variable genetic correlations have been reported between skin wrinkles, skin thickness, body size and wool production, summarized by Williams (1987). Gregory (1982) reported a positive genetic correlation (0.39) between skin thickness and clean fleece weight, but Hynd et al. (1996) observed the opposite, with thinner-skinned sheep pro-

Regulation of Protein Synthesis for Wool Growth

261

ducing more wool than thick-skinned animals. Williams and Morley (1994) reported that sheep that were genetically low wool producers had a higher skin collagen content per unit skin mass than high wool producers.

Hormonal control Hormones affect the mass of active follicular tissue in two different ways. Firstly, hormones influence the follicular cycle of anagen (development), telogen (constant secretion) and catagen (degeneration), which occurs as part of an annual moulting cycle. Secondly, hormones are involved in follicular ‘shut-down’ from nutritional or other stresses. Such shut-down can be distinguished from normal follicle cycles (Schlink and Dollin, 1995). Follicle cycles The impact of annual follicular cycles, coordinated by photoperiod, varies among breeds. The Wiltshire Horn breed may shed the entire fleece, the Romney undergoes seasonal fluctuations of up to 40% in wool growth, while changes in the Merino are relatively limited. Seasonal cycles also appear to affect follicle efficiency (see p. 267). Prolactin mediates photoperiodic effects by synchronizing endogenous follicular cycles, but it does not drive the follicular events. Thus, pharmacological manipulation of prolactin does not affect wool growth (Wallace, 1979), although prolactin receptors are widely distributed in dermal papilla, inner and outer root sheath, germinal matrix, and the sebaceous and sweat glands (Choy et al., 1997). An increased concentration of receptors for insulin-like growth factor-I (IGF-I) has been associated with catagen (Nixon et al., 1997). The effects of the follicular cycle on protein synthesis in skin have not been measured. Follicle shut-down Increased cortisol causes loss of bulb and root sheath cells, resulting in fibre shedding (Chapman and Bassett, 1970). Prolonged treatment with cortisol causes the skin to become thinner, with loss of collagen from the dermis, reduction in the size of sebaceous glands and regression of follicles. Epidermal growth factor can also cause follicle shut-down. Other parts of the glucocorticoid axis, including corticotrophin releasing factor, pro-opiomelanocorticotropin and adrenocorticotropic hormone (ACTH), are found in skin (Slominski et al., 1998), but their involvement in fibre production is unknown. However, local treatment with ACTH can stimulate growth of fibre during the anagen stage in mink (Rose, 1998).

The rate of protein synthesis in skin and follicles The contribution of variation in the rate of protein synthesis in skin to changes in wool growth can be substantial (Table 15.1). However, the proportion of skin protein synthesis that occurs within the follicle is unknown, because the current methods of measurement of protein synthesis rate with isotope tracers treat skin (including follicles) as an integrated organ. Approximately 20% of protein synthesis in skin is secreted as

262

N.R. Adams et al.

wool fibre (Table 15.1), but the proportion of skin protein synthesis accounted for by the follicle must be greater than this. The fibre represents only 10–20% of bulb matrix cells (see p. 266), although these undergo a 13-fold increase in protein mass during keratinization (Short et al., 1968). Additional protein synthesis is associated with the generation, maintenance and maturation of the remaining 80–90% of cells generated in the bulb. Even if these only double in total protein synthesized during maturation and function as inner root sheath, their number is such that they would account for more protein synthesis than that involved in fibre formation. Further protein synthesis occurs in the generation, maturation and maintenance of cells of the outer root sheath, not associated with the bulb (Chapman et al., 1980). Thus, up to half the protein synthesis in sheep skin is likely to be associated with the wool follicle. This value would explain the close relationship between protein synthesis in skin and wool fibre production recorded in Table 15.1.

Dietary supply of amino acids The fractional rate of protein synthesis in the skin, and consequently the rate of wool growth, is highly responsive to changes in feed intake and type of dietary protein (Table 15.3). However, the extent of the increase, and its relativity with increases in other tissues, varies substantially between experimental models (Table 15.3). These differences must be related at least in part to the relative demands of skin and other tissues (see below). Increasing the supply of sulphur amino acids increases the rate of wool growth and the concentration of sulphur amino acids in wool. The wool growth response tends to be higher in sheep with a low overall feed intake (Harris et al., 1994a). Few studies have measured directly the rate of protein synthesis in skin or whole body in response to feeding or infusing sulphur amino acids. Harris et al. (1997) measured an increase in the uptake of cysteine for protein synthesis following cysteine infusion into an isolated skin patch. However, it was not possible to determine whether total protein synthesis increased in skin, or just the ultra-high sulphur protein component. S. Liu (unpublished observations) measured an increase in fractional synthesis rate (FSR) of protein in the skin of 11%, and a significant increase in keratin mRNAs, following methionine infusion into the jugular vein. In this experiment wool growth increased by 37%, indicating that some of the observed increase in wool growth resulted from increased efficiency of incorporation of synthesized protein into wool. Amino acids available for protein synthesis are derived from the diet and also potentially from protein degradation. Protein synthesis for the wool fibre is essentially non-degraded (Harris and Lobley, 1991), but the contribution of other protein degradation in the skin to subsequent protein synthesis is unknown. For example, skin protein content of 46 kg sheep can be estimated as 590 g from MacRae et al. (1993) and at maintenance the amount of protein degradation is 59 g day⫺1 assuming a fractional degradation rate of 10% day⫺1, similar to skin FSR (Rocha et al., 1993). In terms of amino acids, this amount of degraded protein could provide 0.6 g day⫺1 methionine, calculated from amino acid concentrations in skin protein (MacRae et al., 1993). On the other hand, amino acids available to the skin from the circulation can be estimated to be considerably less than this. Using a cardiac output of 4.9 l min⫺1 for a 46 kg sheep,

Regulation of Protein Synthesis for Wool Growth

263

Table 15.3. Fractional synthesis rate (FSR, % day⫺1) of protein in midside skin and skeletal muscle of sheep in response to different diets. Diet or treatment

Muscle FSR

Skin FSR

0.44 M 1.15 M 1.8 M

1.41 2.27 2.93

12.8 18.2 20.6

Merino lambs 24–26 kg, fed 0.83 M

Canola meal Lupin seeds

2.33 2.21

17.3 15.8

D.G. Masters (unpublished observations)a

Merino lambs 25–33 kg

0.6 M 1M

1.64 2.28

15.4 16.9

Liu et al. 1998

⫺ Cimaterol + Cimaterol

1.45 2.76–3.01

11.6 6.3–10.9

0.6 M 1.2 M 1.8 M

1.16 1.67 3.30

Sheep Merino wethers 35–47 kg

Suffolk cross wether, fed 1.7 M 35–41 kg Suffolk cross wether 26–35 kg Romney sheep 42–44 kg Suffolk cross wether 26–35 kg

1.5 M + 2 g cysteine 0.6 M 1.8 M

1.91 2.84

Wholebody FSR 2.20 2.52

Source N.R. Adams (unpublished observations)

Nash et al. (1994) 2.83 3.76 4.73

Harris et al. (1992)b

12.8 22.8

Harris et al. (1994b)c

5.6 9.5

Lobley et al. (1992)

aSource

of protein was lupin seed or canolameal. in muscle was measured using arterio-venous technique. All the others were measured using flooding-dose techniques. Isotope-labelled phenylalanine was used as the tracer. cFSR in skin was measured from the uptake of [35S]cysteine by skin using arterio-venous technique.

bFSR

a proportion of 0.06 to the whole skin, and concentrations in blood of 20 ␮M for methionine (Harris et al., 1989, 1997), calculations indicate a supply of 1.3 g day⫺1 methionine. However, only 10–35% is taken up from the blood (Harris et al., 1997), so the effective amino acid supply to the skin from the blood is considerably less than the total amount available to skin from protein degradation in situ. Despite these calculations, evidence that protein degradation makes a significant contribution to wool growth is still weak. For example, lines of sheep that differ in weaning weight differed primarily in protein degradation rate (Oddy et al., 1995), yet wool growth in these lines was similar (Pattie and Williams, 1967).

264

N.R. Adams et al.

Competition for amino acids Growth At high nutrient intake, Graham and Searle (1972) found a constant relationship between wool production and feed intake from birth, indicating that wool was similar to other body tissues in the age-associated reduction in the efficiency of incorporation of amino acids into protein. However, at lower feed intakes Doyle and Egan (1983) found that body tissues competed more strongly than wool for nutrients in young sheep, while in older sheep wool growth was maintained at the expense of body tissues. Further evidence of competition for nutrients between wool growth and other body tissues was obtained by Graham and Searle (1982) and Cronje and Smuts (1994). While the FSR of protein responds to changes in dietary intake in a number of tissues, including skin, muscle and gut (Lobley et al., 1992, 1994), the magnitude of the response is not always consistent between tissues. As shown in Table 15.3, increasing feed intake from 0.6 M to 1.8 M in the Suffolk-cross lambs resulted in large increases in skin FSR (68%) but less in muscle (29%), while the opposite was observed in the Merino, with a change in feeding from 0.6 M to 1.6 M increasing skin FSR by 10% and muscle FSR by 60% (Liu et al., 1998). Furthermore, skin FSR was reported to be three times that of muscle in Suffolk-cross lambs (Lobley et al., 1992) and six times that of muscle in Merino lambs (Liu et al., 1998). These differences may be related to the high rates of wool growth and less muscle growth in Merino sheep, but other factors such as the growth demand by the animal and the length of re-feeding period must also be important. Pregnancy and lactation During pregnancy there is usually a specific reduction in the proportion of whole body protein deposited in wool, relative to maternal and fetal tissue. Merino ewes fed to maintain maternal liveweight increased non-wool body protein in the ewe and fetus by 6% between days 127 and 144 of pregnancy. During the same period, wool growth decreased by 6% (Masters and Mata, 1996). In lactation, the situation is reversed with maintenance or even an increase in wool protein production at the same time as a reduction in maternal body protein (Masters and Stewart, 1990; Masters and Mata, 1996). These findings are consistent with the hormonal milieu increasing the priority of fetal over maternal tissues during pregnancy and lactation. Measurements of protein synthesis in skin and muscle during the final month of pregnancy indicated that reduced wool growth resulted from reduced skin mass (skin thickness) and a slight but variable reduction in FSR per unit of skin relative to nonpregnant sheep of the same maternal weight (Table 15.1; S. Liu, D.G. Masters and H. O’Donohoe, unpublished observations). The situation during lactation remains to be explored. Hormonal effects Effects of hormones on protein synthesis have been well reviewed (MacRae and Lobley, 1991; Bell et al., 1998) and are summarized in Table 15.4. In contrast to most other areas of animal production, there are no hormones with a homeorrhoetic function to direct nutrients to the skin. However, hormones do affect wool growth through their effects on whole body protein synthesis, or by affecting protein metabolism in other

Regulation of Protein Synthesis for Wool Growth

265

Table 15.4. Summary of increases (≠), decreases (Ø) or no significant change (Æ) in whole body protein metabolism and in wool growth, reported to result from treatment with hormones.

Hormone

Protein synthesis

Protein degradation

Specific follicle effects

Wool effect

Thyroxine Androgens ␤-adrenergic agonist Growth hormone Insulin IGF-I Cortisol Prolactin

≠ ≠? ≠ ≠ Æ Æ, ≠? Æ Æ

≠ Ø Ø Æ Ø Ø

Follicle efficiency Nil Nil Nil Nil Nil Follicle shut-down Follicle efficiency

≠or Æ ≠ Ø ≠ or Ø Æ Æ or ≠ Ø Ø or Æ

≠ Æ

organs. In addition to these effects, specific effects of hormones on the mass or efficiency of wool follicles are dealt with on pp. 261 and 264. Most of the effects of hormones on wool growth are consistent with their effects on general protein metabolism. For example, the small increase in wool growth reported in wethers treated with androgens (Southcott and Royal, 1971; Hynd and James, 1987) probably resulted from an increase in feed intake. Insulin and cortisol, which affect protein degradation rather than synthesis, have no effect on wool growth under normal physiological concentrations. High concentrations of cortisol may have specific effects on follicular function (p. 261). Growth hormone may increase (Johnsson et al., 1985) or decrease (Wynn et al., 1988) wool growth, depending on the relative balance between an increase in overall body protein synthesis resulting from increased feed intake, and the diversion of nutrients to tissues more responsive to growth hormone, such as gut and muscle. Diversion of nutrients is even more marked with the adrenergic ␤-agonist cimaterol, which decreased wool growth by 16% and shorn skin weight by 9%, while increasing carcass protein and decreasing lipid deposition (Fennessy et al., 1990). Nash et al. (1994) found that the reduction in wool growth was directly counter-balanced by the increase in protein deposition in muscle of treated sheep. Recent work has explored the effects of IGF-I on wool growth. Expression of IGFI with a keratin promoter in transgenic sheep increased wool growth during spring and summer, but not in winter, so that staple strength was decreased (Damak et al., 1996). More extensive studies (Su et al., 1998) on the subsequent generation of these transgenic animals were unable to detect a significant increase in wool growth or in plasma IGF-I. Sheep treated with IGF-I had a transient increase in skin protein synthesis which disappeared by 24 h (Lobley et al., 1998). The extensive buffering of IGF-I through specific binding proteins makes assessment of its role difficult. It mediates the stimulation of protein synthesis by growth hormone (Bell et al., 1998) and may affect the interaction between cortisol and low nutritional status (Chapman and Bassett, 1970), because treatment with IGF-I can prevent the deleterious effects of glucocorticoids on nitrogen metabolism (Tomas, 1998).

266

N.R. Adams et al.

Responsiveness to nutrient supply Merino sheep with a high genetic capacity for wool growth respond more to nutrition, although the magnitude of the response varies among genotypes (Jackson and Roberts, 1970; Woolaston, 1987). Differences in responsiveness probably reflect, in part, differences in overall protein metabolism. For example, the responsiveness of protein synthesis rate to nutrition is reduced in both skin and muscle of sheep selected for high staple strength (Fig. 15.2). These sheep also differed in nutrient partitioning between muscle and wool, such that the sheep with a less volatile wool growth response to nutrition had greater changes in liveweight (Adams et al., 1997). A similar apparent competition for protein between wool and muscle was reported by Cronje and Smuts (1994). Follicle-specific mechanisms may also affect the responsiveness to nutrient supply. For example, Harris et al. (1994b) suggested that the rate of uptake of cyst(e)ine by follicle cells may be genetically determined. Fibre length and diameter can respond differently to nutrition in different flocks (Jackson and Roberts, 1970), depending on factors such as cortical cell length and the dimensions of the dermal papilla (Hynd, 1994a). The relative importance of specific follicle effects and whole body mechanisms, in determining the effects of nutrition on wool growth, remains unknown.

The proportion of skin protein synthesis in wool The remaining factor that may affect wool growth is the proportion of skin protein synthesis that becomes fibre. Although this is similar across genotypes (Table 15.1), unpublished work indicates that efficiency differences of 24% between high and low fleece weight genotypes contributed to a total difference of 53% in wool growth rate (S. Liu, D.G. Masters and M.J. Nancarrow, unpublished observations). At a different level,

Fig. 15.2. Fractional synthesis rate (FSR) of protein in skin (a) and muscle (b) from ewes selected for low (䢇) and high (嘷) staple strength. Significant (P < 0.05) interaction between genotype and feed intake for both tissues.

Regulation of Protein Synthesis for Wool Growth

267

the phenomenon has been studied as the proportion of cells from the bulb matrix that enter the fibre (as opposed to the root sheath). Normally, between 9 and 40% of cells formed in the bulb enter the fibre, the remaining cells becoming root sheath and subsequently lost (Wilson and Short, 1979; Hynd, 1989). This factor has been called the ‘follicle efficiency’. Its exact relationship to protein synthesis is not defined.

Nutrition and genotype Short et al. (1968) found that a dramatic increase in nutrition, which increased wool production by 250%, only increased the average proportion of cells entering the fibre from 14.6% to 18.3%. Hynd (1989) concluded that follicle efficiency was independent of nutrition. Both Williams and Winston (1987) and Hynd (1989) concluded that follicle efficiency contributes about 10% to genetic differences between sheep in wool production. There was a general trend to higher follicle efficiency in sheep with a higher fibre diameter that produced more wool (Hocking Edwards and Hynd, 1992).

Hormones In addition to its role in mediating the effect of photoperiod on the follicular cycle described on p. 261, prolactin may also mediate seasonal changes in the efficiency of wool growth. Holle et al. (1994) reported that seasonal changes in wool growth were accompanied by changes to the relative distribution of cells between the fibre and inner root sheath. This was estimated by the ‘production ratio’, the ratio of area of fibre to root sheath in horizontal sections of skin (Butler and Wilkinson, 1979). However, Hynd (1989) found a poor correlation between the production ratio and follicle efficiency. Measures of seasonal changes in the ratio of protein synthesis to wool production might offer valuable insights. Hynd (1994b) showed that abnormally low thyroxine concentrations reduced follicle efficiency by 40%, and also reduced the production ratio. Cell division was only reduced by 16%, although the proportion of inactive follicles was increased. Thus, it appears that thyroxine affects follicle efficiency more than bulb activity. However, it is likely that such low levels of thyroxine do not occur in normal husbandry.

Conclusions The impact of genotype, nutrition and physiological state on wool growth can be integrated with other bodily functions by considering it as part of the overall protein synthesis in the body. Indeed, the rate of protein synthesis for wool growth broadly reflects that in other tissues, so wool growth offers a simple index of changes in protein synthesis rate in the body. Furthermore, protein synthesis and protein degradation rates together provide a sounder basis for understanding of nutrient partitioning between wool and other tissues than do concepts based on mass flow. The rate of wool growth is affected both by specific follicle mechanisms and by factors controlling protein metabolism in the whole body, but their relative importance

268

N.R. Adams et al.

is unclear in most situations. Although specific morphological characteristics of follicles correlate closely with wool growth (e.g. Jackson et al., 1975), some of these may simply reflect systemic effects, such as nutrient supply. The relative importance of local and systemic factors remains a major challenge to wool scientists. Much of the information needed to estimate wool growth as a part of protein synthesis is still sketchy. There are limited data on protein synthesis rates in the skin, and none that distinguish the follicle from the remaining skin. Even simple characteristics, such as the total fraction of skin volume made up by germinative epithelium, are poorly documented. However, as the missing data become available and models of protein synthesis become more widely used, sheep farmers and animal scientists will be able to apply a greater mechanistic understanding to enable more accurate and robust prediction of the outcomes of breeding and feeding.

Acknowledgements Research by CSIRO Animal Production and the Co-operative Research Centre for Premium Quality Wool is supported by Australian wool growers through the Australian Wool Research and Promotion Organization.

References Adams, N.R., Briegel, J.R. and Ritchie, A.J.M. (1997) Wool and liveweight responses to nutrition by Merino sheep genetically selected for high or low staple strength. Australian Journal of Agricultural Research 48, 1129–1137. Allden, W.G. (1979) Feed intake, diet composition and wool growth. In: Black, J.L. and Reis, P.J. (eds) Physiological and Environmental Limitations to Wool Growth. The University of New England Publishing Unit, Armidale, New South Wales, pp. 61–78. Bell, A.W., Bauman, D.E., Beerman, D.H. and Harrell, R.J. (1998) Nutrition, development and efficiency of growth modifiers in livestock species. Journal of Nutrition 128, 360S–363S. Black, J.L. and Reis, P.J. (1979) Speculation on the control of nutrient partition between wool growth and other body functions. In: Black, J.L. and Reis, P.J. (eds) Physiological and Environmental Limitations to Wool Growth. The University of New England Publishing Unit, Armidale, New South Wales, pp. 269–293. Black, J.L., Robards, G.E. and Thomas, R. (1973) Effects of protein and energy intake on the wool growth of merino wethers. Australian Journal of Agricultural Research 24, 399–412. Bowman, P.J., Cottle, D.J., White, D.H. and Bywater, A.C. (1993) Simulation of wool growth rate and fleece characteristics of merino sheep in southern Australia. Part 1. Model description. Agricultural Systems 43, 287–299. Butler-Hogg, B.W. (1984) Growth patterns in sheep: wool growth during weight loss and subsequent compensatory growth. Journal of Agricultural Science, Cambridge 102, 105–109. Butler, L.G. and Wilkinson, B.R. (1979) Efficiency of wool production and the mechanics of wool follicle production. New Zealand Journal of Agricultural Research 22, 543–545. Chapman, R.E. and Bassett, J.M. (1970) The effects of prolonged administration of cortisol on the skin of sheep on different planes of nutrition. Journal of Endocrinology 48, 649–663. Chapman, R.E., Downes, A.M. and Wilson, P.A. (1980) Migration and keratinization of cells in wool follicles. Australian Journal of Biological Science 33, 587–603. Choy, V.J., Nixon, A.J. and Pearson, A.J. (1997) Distribution of prolactin receptor immunoreac-

Regulation of Protein Synthesis for Wool Growth

269

tivity in ovine skin and changes during the wool follicle growth cycle. Journal of Endocrinology 155, 265–275. Cronjé, P.B. and Smuts, M. (1994) Nutrient partitioning in Merino rams with different wool growth rates. Animal Production 59, 55–60. Damak, S., Su, H.-Y., Jay, N.P. and Bullock, D.W. (1996) Improved wool production in transgenic sheep expressing insulin-like growth factor 1. Biotechnology 14, 185–188. Doyle, P.T. and Egan, J.K. (1983) The utilization of nitrogen and sulfur by weaner and mature sheep. Australian Journal of Agricultural Research 34, 433–439. Fennessy, P.F., McEwan, J.C., Lord, E.A., Greer, G.J., Bain, W.E., Johnstone, P.D., Knowler, M.A., Dalrymple, R.H. and Ingle, D.L. (1990) Effect of Cimaterol implants on lamb growth and carcass traits. New Zealand Journal of Agricultural Research 33, 413–427. Freer, M., Moore, A.D. and Donnelly, J.R. (1997) Decision support systems for Australian grazing enterprises. II The animal biology model for feed intake, production and reproduction and the Grazfeed DSS. Agricultural Systems 53, 77–126. Graham, N.M. and Searle, T.W. (1972) Growth in sheep. II Efficiency of energy and nitrogen utilization from birth to 2 years. Journal of Agricultural Science, Cambridge 79, 383–389. Graham, N.M. and Searle, T.W. (1982) Energy and nitrogen utilisation for body growth in young sheep from two breeds with differing capacities for wool growth. Australian Journal of Agricultural Research 33, 607–615. Gregory, I.P. (1982) Genetic studies of South Australian Merino sheep. IV genetic, phenotypic and environmental correlations between various wool and body traits. Australian Journal of Agricultural Research 33, 363–373. Hales, J.R.S. and Fawcett, A.A. (1993) Wool production and blood supply to skin and other tissues in sheep. Journal of Animal Science 71, 492–498. Harris, P.M. and Lobley, G.E. (1991) Amino acid and energy metabolism in the peripheral tissues of ruminants. In: Tsude, T., Sasaki, Y. and Kawashima, R. (eds) Physiological Aspects of Digestion and Metabolism in Ruminants. Academic Press, San Diego, California, pp. 201–230. Harris, P.M., Dellow, D.W. and Sinclair, B.R. (1989) Preliminary ‘in vivo’ measurements of protein and energy metabolism in the skin of sheep. Australian Journal of Agricultural Research 40, 879–888. Harris, P.M., Skene, P.A., Buchan, V., Milne, E., Calder, A.G., Anderson, S.E., Connell, A. and Lobley, G.E. (1992) Effect of food intake on hind-limb and whole-body protein metabolism in young growing sheep: chronic studies based on arterio-venous techniques. Species comparisons of tissue protein metabolism: effects of age and hormonal metabolism. British Journal of Nutrition 68, 389–407. Harris, P.M., Lee, J., Sinclair, B.R. and Treloar, B.P. (1994a) The effect of whole body cysteine supplementation on cysteine utilization by the skin of a well-fed sheep. Proceedings of the New Zealand Society of Animal Production 54, 139–142. Harris, P.M., Lee, J., Sinclair, B.R., Treloar, B.P. and Gurnsey, M.P. (1994b) Effect of food intake on energy and protein metabolism in the skin of Romney sheep. British Journal of Nutrition 71, 647–660. Harris, P.M., Sinclair, B.R., Treloar, B.P. and Lee, G.J. (1997) Short-term changes in whole body and skin sulfur amino acid metabolism of sheep in response to supplementary cysteine. Australian Journal of Agricultural Research 48, 137–146. Hocking Edwards, J.E. and Hynd, P.I. (1992) Cellular characteristics of wool follicles and fibres in finewool and strongwool Merinos. Australian Journal of Agricultural Research 43, 355–365 (Abstract). Holle, S.A., Harris, P.M., Davies, A.S. and Birtles, M.J. (1994) Wool follicle morphology and cell proliferation in New Zealand Romney sheep; a seasonal comparison of fleeceweightselected and control rams. Australian Journal of Agricultural Research 45, 769–782.

270

N.R. Adams et al. Hynd, P.I. (1989) Effects of nutrition on wool follicle cell kinetics in sheep differing in efficiency of wool production. Australian Journal of Agricultural Research 40, 409–417. Hynd, P.I. (1994a) Follicular determinants of the length and diameter of wool fibres. I. Comparison of sheep differing in fibre length/diameter ratio at two levels of nutrition. Australian Journal of Agricultural Research 45, 1137–1147. Hynd, P.I. (1994b) Follicular determinants of the length and diameter of wool fibres. II. Comparison of sheep differing in thyroid hormone status. Australian Journal of Agricultural Research 45, 1149–1157. Hynd, P.I. and James, R.E. (1987) The effect of trenbolone acetate and trenbolone acetate plus oestradiol-17␤ on wool growth. Journal of Agricultural Science, Cambridge 108, 501–503. Hynd, P.I., Ponzoni, R.W., Grimson, R.J., Jaensch, K.S., Smith, D. and Kenyon, R.V. (1996) Wool follicle and skin characters – their potential to improve wool production and quality in Merino sheep. Wool Technology and Sheep Breeding 44, 167–177. Jackson, N. and Roberts, E.M. (1970) Comparison of three Australian Merino strains for wool and body traits. I. Genetic means of studs and strains and their interactions with years and sexes. Australian Journal of Agricultural Research 21, 815–835. Jackson, N., Nay, T. and Turner, H.N. (1975) Response to selection in Australian Merino sheep. VII Phenotypic and genetic parameters for some wool follicle characteristics and their correlation with wool and body traits. Australian Journal of Agricultural Research 26, 937–957. Johnsson, I.D., Hart, I.C. and Butler-Hogg, B.W. (1985) The effects of exogenous bovine growth hormone and bromocryptine on growth, body development, fleece weight and plasma concentrations of growth hormone, insulin and prolactin in female lambs. Animal Production 41, 207–217. Knap, P.W. and Schrama, J.W. (1996) Simulation of growth in pigs: approximation of protein turnover characteristics. Animal Science 63, 533–547. Liu, S., Mata, G., O’Donoghue, H. and Masters, D.G. (1998) The influence of live weight, liveweight change and diet on protein synthesis in the skin and skeletal muscle in young Merino sheep. British Journal of Nutrition 79, 267–274. Lobley, G.E., Harris, P.M., Skene, P.A., Brown, D., Milne, E., Calder, A.G., Anderson, S.E., Garlick, P.J., Nevison, I. and Connell, A. (1992) Response in tissue protein synthesis to sub- and supra-maintenance intake in young growing sheep: comparison of large-dose and continuous-infusion techniques. British Journal of Nutrition 68, 373–388. Lobley, G.E., Connell, A., Milne, E., Newman, A. and Ewing, T.A. (1994) Protein synthesis in splanchnic tissues of sheep offered two levels of intake. British Journal of Nutrition 71, 3–12. Lobley, G.E., Lee, J., Hocking Edwards, J.E. and Harris, P.M. (1998) A comparison of changes in whole body and skin amino acid metabolism of sheep in response to 24 h continuous infusions of variants of insulin-like growth factor 1. Canadian Journal of Animal Science 77, 695–706. MacRae, J.C. and Lobley, G.E. (1991) Physiological and metabolic implications of conventional and novel methods for the manipulation of growth and production. Livestock Production Science 27, 43–59. MacRae, J.C., Walker, A., Brown, D. and Lobley, G.E. (1993) Accretion of total protein and individual amino acids by organs and tissues of growing lambs and the ability of nitrogen balance techniques to quantitate protein retention. Animal Production 57, 237–245. Masters, D.G. and Mata, G. (1996) Responses to feeding canola meal or lupin seed to pregnant, lactating, and dry ewes. Australian Journal of Agricultural Research 47, 1291–1303. Masters, D.G. and Stewart, C.A. (1990) Wool growth and reproduction. In: Oldham, C.M., Martin, G.B. and Purvis, I.W. (eds) Reproductive Physiology of Merino Sheep. Concepts and Consequences. School of Agriculture, The University of Western Australia, Perth, pp. 265–274. Mata, G. and Masters, D.G. (1999) Rumen protected methionine improves the efficiency of use

Regulation of Protein Synthesis for Wool Growth

271

of metabolisable protein for wool growth. In: Stone, G., Forbes, T.D.A., Stith, J.W. and Byers, F.M. (eds) Proceedings of the 5th International Symposium on the Nutrition of Herbivores, San Antonio Texas. Online poster http://cnrit.tamu.edu/conf/isnh/post-online/post 0045. Murray, D.M. and Slezacek, O. (1994) Effects of a 50-day period of slow liveweight loss on the body composition of crossbred wethers. Proceedings of the Australian Society for Animal Production 20, 397. Nagorka, B.N. (1977) The description and analysis of wool growth. Australian Journal of Agricultural Research 28, 737–746. Nash, J.E., Rocha, H.J.G., Buchan, V., Calder, G.A., Milne, E., Quirke, J.F. and Lobley, G.E. (1994) The effect of acute and chronic administration of the ␤-agonist, cimaterol, on protein synthesis in ovine skin and muscle. British Journal of Nutrition 71, 501–513. Nixon, A.J., Ford, C.A., Oldham, J.M. and Pearson, A.J. (1997) Localisation of insulin-like growth factor receptors in skin follicles of sheep (Ovis aries) and changes during an induced growth cycle. Comparative Biochemistry and Physiology 118A, 1247–1257. Oddy, V.H., Speck, P.A., Warren, H.M. and Wynn, P.C. (1995) Protein metabolism in lambs from lines divergently selected for weaning weight. Journal of Agricultural Science, Cambridge 124, 129–137. Pattie, W.A. and Williams, A.J. (1967) Selection for weaning weight in Merino sheep. 3. Maintenance requirements and the efficiency of conversion of feed to wool in mature ewes. Australian Journal of Experimental Agriculture and Animal Husbandry 7, 117–125. Rocha, H.J.G., Nash, J.E., Connell, A. and Lobley, G.E. (1993) Protein synthesis in ovine muscle and skin: sequential measurements with three different amino acids based on the largedose procedure. Comparative Biochemistry and Physiology 105B, 301–307. Rose, J. (1998) Adrenocorticotropic hormone (ACTH) but not alpha-melanocyte stimulating hormone (a-MSH) as a mediator of adrenalectomy induced hair growth in mink. Journal of Investigative Dermatology 110, 456–457. Sainz, R.D. and Wolff, J.E. (1990) Development of a dynamic, mechanistic model of lamb metabolism and growth. Animal Production 51, 535–549. Schlink, A.C. and Dollin, A.E. (1995) Abnormal shedding contributes to the reduced staple strength of tender wool in Western Australian Merinos. Wool Technology and Sheep Breeding 43, 268–284. Scobie, D.R. and Woods, J.L. (1992) A review implicating a two compartment model for the processes of cell division and differentiation in the wool follicle. Proceedings of the New Zealand Society of Animal Production 52, 265–271. Short, B.F., Wilson, P.A. and Schinckel, P.G. (1968) Proliferation of follicle matrix cells in relation to wool growth. In: Lyne, H.G. and Short, B.F. (eds) Biology of Skin and Hair Growth. Angus and Robertson, Sydney, pp. 409–426. Slominski, A., Ermak, G., Mazurkiewicz, J.E., Baker, J. and Wortsman, J. (1998) Characterization of corticotropin-releasing hormone (CRH) in human skin. Journal of Clinical Endocrinology and Metabolism 83, 1020–1024. Southcott, W.H. and Royal, W.M. (1971) Effect of implanted testosterone propionate on the growth and wool production of merino wethers. Australian Journal of Agricultural Research 22, 271–282. Su, H.Y., Jay, N.P., Gourley, T.S., Kay, G.W. and Damak, S. (1998) Wool production in transgenic sheep – results from first-generation adults and second-generation lambs. Animal Biotechnology 9, 135–147. Tomas, F.M. (1998) The anti-catabolic efficiency of insulin-like growth factor-1 is enhanced by its early administration to rats receiving dexamethasone. Journal of Endocrinology 157, 89–97. Wallace, A.L.C. (1979) The effect of hormones on wool growth. In: Black, J.L. and Reis, P.J. (eds) Physiological and Environmental Limitations to Wool Growth. University of New England Publishing Unit, Armidale, New South Wales, pp. 257–268.

272

N.R. Adams et al. Williams, A.J. (1987) Physiological consequences of selection for increased fleece weight. In: McGuirk, B.J. (ed.) Merino Improvement Programs in Australia. Australian Wool Corporation, Melbourne, pp. 481–494. Williams, A.J. and Morley, F.C. (1994) Influence of dietary intake and genetic capacity for wool growth on the composition of mid-trunk skin of Merino sheep. Australian Journal of Agricultural Research 45, 1715–1729. Williams, A.J. and Winston, R.J. (1987) A study of the characteristics of wool follicle and fibre in merino sheep genetically different in wool production. Australian Journal of Agricultural Research 38, 743–755. Wilson, P.A. and Short, B.F. (1979) Cell proliferation and cortical cell production in relation to wool growth. Australian Journal of Biological Science 32, 317–327. Woolaston, R.R. (1987) Genotype ⫻ environment interactions and their possible impact on breeding programs. In: McGuirk, B.J. (ed.) Merino Improvement Programs in Australia. Australian Wool Corporation, Melbourne, pp. 421–435. Wynn, P.C., Wallace, A.L.C., Kirby, A.C. and Annison, E.F. (1988) Effects of growth hormone administration on wool growth in merino sheep. Australian Journal of Biological Science 41, 177–187.

VI

Reproduction, Pregnancy and Lactation

16

Regulation of Macronutrient Partitioning between Maternal and Conceptus Tissues in the Pregnant Ruminant

A.W. BELL AND R.A. EHRHARDT Department of Animal Science, Cornell University, Ithaca, New York, USA

Introduction The concept that pregnant animals partition available nutrients to favour their developing offspring was formally postulated by Hammond (1947) who considered that tissues compete for circulating nutrients on the basis of their relative metabolic rates. This idea was reinforced by early demonstrations of high metabolic rates in tissues of the conceptus relative to those of the dam (Meschia et al., 1980). However, more recent thinking on the extracellular regulation of nutrient partitioning has focused on endocrine coordination rather than tissue competition as a general explanatory mechanism, building on the concept of homeorhesis elaborated by Bauman and Currie (1980). This concept more satisfactorily accommodates the coordinated metabolic compromises implicit in the constraint of fetal growth during late pregnancy, to levels that optimize opportunity for neonatal survival and postnatal development, yet minimize excessive depletion of maternal energy and protein reserves during late pregnancy. This chapter will describe patterns of partitioning of macronutrients between maternal and conceptus tissues, as a background to more detailed consideration of their mechanistic explanations involving placental functions and adaptive responses of maternal (non-uterine) tissues to pregnancy.

Patterns of nutrient utilization and partitioning during pregnancy Fetal nutrient requirements The conceptus (fetus(es), placenta, associated fetal membranes, and supporting uterine tissues) makes extensive, direct demands upon maternal supplies of glucose and amino acids, especially during late pregnancy. Numerous in vivo studies of conceptus metabolism in sheep have focused mostly on fetal metabolism in terms of umbilical exchanges © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

275

276

A.W. Bell and R.A. Ehrhardt

of oxygen, nutrients and metabolites (see Battaglia and Meschia, 1988; Bell, 1993). These and similar data from pregnant cows (Comline and Silver, 1976; Reynolds et al., 1986; Ferrell, 1991) have been used to construct metabolic balance sheets for fetal sheep (Bell, 1993) and cattle (Bell, 1995). In both species during late gestation, 35–40% of fetal energy is supplied as glucose and its fetal-placental metabolite, lactate, and a further 55% is taken up as amino acids. Most of the residual 5–10% is contributed by acetate, placental transfer of which is meagre relative to its abundance and energetic importance in the maternal system. Placental capacity for maternal–fetal transport of long-chain, non-esterified fatty acids (NEFAs) and ketoacids is even more limited (see Bell, 1993), denying fetal access to substrates derived from maternal fat mobilization and thus constraining fetal ability to grow at the direct expense of maternal energy reserves. Almost all of the nitrogen acquired by the fetus is in the form of amino acids, although a small net umbilical uptake of ammonia, derived from placental deamination of amino acids, is detectable in fetal sheep during mid (Bell et al., 1989a) and late gestation (Holzman et al., 1977). Patterns of utilization of glucose and amino acids for oxidative metabolism and growth in fetal ruminants are reviewed elsewhere (Battaglia and Meschia, 1988; Bell, 1993, 1995; Hay, 1996, 1998).

Metabolism of non-fetal conceptus tissues The major contribution of the non-fetal components of the gravid uterus, especially the placenta, to oxygen and nutrient requirements of the conceptus is sometimes ignored. However, these requirements greatly affect the partitioning of nutrients within the gravid uterus and add substantially to the nutrient demands upon the dam. In latepregnant ewes and cows, the aggregate weight of placentomes, consisting of fetal (cotyledonary) and maternal (caruncular) tissues, is less than 15% of that of the attached fetus. However, the weight-specific metabolic rate of the placenta is so great that the uteroplacental tissues (placentomes, endometrium, myometrium) consume 35–50% of the oxygen and 60–70% of the glucose taken up by the uterus in ewes (Meschia et al., 1980) and cows (Reynolds et al., 1986). Net uteroplacental consumption of amino acids is lower relative to fetal uptake, because growth of the placenta and uterine tissues is negligible in sheep (Ehrhardt and Bell, 1995) and small in cattle (Bell et al., 1995) during late pregnancy. Nevertheless, net removal by the uteroplacental tissues has been estimated to account for 24% of uterine uptake of amino acid nitrogen in well-fed ewes during late pregnancy (Chung et al., 1998).

Partitioning of glucose and amino acids between conceptus and maternal tissues In well-fed, monotocous ewes during late pregnancy, uterine uptake of glucose consumes 30–50% of maternal glucose supply (Prior and Christenson, 1978; Hay et al., 1983b; Oddy et al., 1985; Leury et al., 1990), which accounts for all of the pregnancyinduced increment in whole-body production and utilization of glucose (Petterson et al., 1993; Freetly and Ferrell, 1998). About 10% of this glucose carbon is returned to the maternal circulation as lactate (Meschia et al., 1980; Faichney et al., 1981), presum-

Regulation of Macronutrient Partitioning

277

ably resulting from glycolysis in uteroplacental tissues of maternal origin, and contributing to the increased cycling between glucose and lactate that is characteristic of ruminants during late pregnancy (van der Walt et al., 1983). No studies have directly addressed the partitioning of amino acids between the gravid uterus and maternal non-uterine tissues. However, we have used traditional nitrogen balance and comparative slaughter techniques to estimate the partition of apparently digested crude protein in ditocous ewes that were fed to predicted energy and protein requirements between days 110 and 140 of pregnancy (McNeill et al., 1997). Calculations suggest that as much as 80% of apparently digested crude protein was partitioned to the gravid uterus, the remainder being used to support increased metabolism and net deposition of amino acids in the developing mammary glands and visceral organs (Bell and Ehrhardt, 1998). However, it is notable that even though the ewes were fed to predicted requirement for dietary protein (> 90 g day⫺1 digestible crude protein), the available pool of circulating amino acids was augmented by the net mobilization of protein from maternal carcass tissues (mostly skeletal muscle), amounting to almost 10% of the digestible crude protein intake.

Modifying effects of maternal nutrition and placental size Direct effects of maternal nutrition Partitioning of glucose between conceptus and maternal tissues during late pregnancy is affected by maternal plane of nutrition (see Bell, 1993; Bell and Ehrhardt, 1998). Previously well-conditioned, ditocous ewes fed at 60% of predicted energy requirement for 3 weeks were able to maintain absolute rates of uterine uptake of glucose despite a 25% reduction in maternal glucose entry rate. Fetal growth was unimpaired but maternal loss of body weight and condition was appreciable. In contrast, more severe undernutrition (30–40% energy requirement for 2–3 weeks) or fasting (5 days) caused a marked decrease in uterine uptake of glucose that was proportional to that in maternal glucose production. Possible mechanisms for these phenomena, involving adaptations in the placenta and maternal tissues, are discussed in later sections. Effects of maternal nutrition on partitioning of amino acids to the conceptus have been little studied. In ewes fasted for 5 days, umbilical net uptake of amino acids was relatively unaffected despite reduced maternal blood concentrations of most amino acids (Lemons and Schreiner, 1983; Liechty et al., 1991). This implies that potentially negative effects of severe, short-term nutrient deprivation on fetal access to maternal amino acids were offset by maternal mobilization of tissue protein reserves and active transport of amino acids by the placenta. Even if these putative compensatory mechanisms are successful in sustaining fetal amino acid supply, a decrease in fetal tissue protein synthesis and growth is to be expected because of increased fetal catabolism of amino acids to substitute for lack of glucose as an energy source (see Bell, 1993). Specific effects of protein deprivation on partitioning of amino acids between maternal tissues and the conceptus have not been examined in pregnant ruminants. However, in ditocous ewes fed a protein-deficient (80 g crude protein kg⫺1 dry matter) but energy-sufficient diet for 30 days in late pregnancy, uterine (mostly fetal) nitrogen accretion was reduced by 20% compared with that in ewes fed to predicted

278

A.W. Bell and R.A. Ehrhardt

requirement (120 g kg⫺1 crude protein) (McNeill et al., 1997). This occurred despite a substantial net loss of protein from maternal carcass tissues and reduced net accretion of protein in maternal visceral and mammary tissues, indicating a major shift in partitioning of maternal amino acids to partially offset the effects of dietary deficiency on conceptus growth (Fig. 16.1). Indirect effects of maternal nutrition Effects of maternal nutrition in early and mid pregnancy may have later effects on nutrient partitioning to the conceptus via influences on growth and highly correlated indices of functional capacity of the placenta. Moderate undernutrition of ewes during early to mid pregnancy, when placental growth is rapid (Ehrhardt and Bell, 1995), has caused conflicting positive (Faichney and White, 1987; McCrabb et al., 1992) and negative (McCrabb et al., 1992; Clarke et al., 1998) effects on placental size. Variation in body condition during early pregnancy appears to at least partly explain this confusion, in that fatter ewes responded to underfeeding with an increase in placental size, whereas the opposite occurred in lean ewes (McCrabb et al., 1992). This suggests that if maternal energy stores are available, the dam will mobilize them for her own use, permitting a compensatory response in placental growth.

Fig. 16.1. Protein deposition between days 110 and 140 of pregnancy in maternal tissue components of ditocous ewes fed diets containing different levels of dietary crude protein. Histograms are means for eight ewes. Pooled standard errors were 214 g for carcass, 84 g for organs, and 44 g for mammary gland. Within tissue components, means with different letters are significantly different (P < 0.05). Adapted from the data of McNeill et al. (1997).

Regulation of Macronutrient Partitioning

279

In distinct contrast, recent novel studies have shown that overfeeding and rapid growth of primiparous ewes during early–mid pregnancy causes profound reductions in placental growth, followed by severe fetal growth retardation with all the hallmarks of placental insufficiency (Wallace et al., 1996, 1999). Intriguingly, the negative effects of overfeeding before day 50 of pregnancy were ameliorated by feed restriction after day 50; conversely, overfeeding after day 50 partially induced placental and fetal growth retardation in ewes fed at a moderate level up to that time (Table 16.1; Wallace et al., 1999). Maternal body condition may also affect the partitioning of nutrients during late pregnancy by mechanisms not directly related to placental size or function. For example, McNeill et al. (1998) observed that when given ad libitum access to feed during late pregnancy, lean ewes ate more than fatter ewes but partitioned the extra nutrients to maternal rather than conceptus tissues. Also, fatter ewes were better able than lean ewes to sustain conceptus growth during moderate undernutrition in late pregnancy (McNeill et al., 1999). Placental size did not differ appreciably between lean and fatter ewes in either of these studies, and was not considered to be an important determinant of differences in nutrient partitioning between maternal and conceptus tissues. Effects of non-nutritional influences on placental size Environmental heat stress or pre-mating carunclectomy can cause dramatic reductions in placental size and functional capacity (see Bell, 1987; Robinson et al., 1995), and might, therefore, be expected to influence the partitioning of nutrients between maternal and conceptus tissues. Plasma glucose concentrations were elevated in chronically heat-treated (Bell et al., 1989b) and carunclectomized ewes (Falconer et al., 1985) fed at the same levels as controls during late pregnancy, suggesting that placental restriction may have caused a repartitioning of maternal glucose away from the conceptus. Unfortunately, the only experiment designed to test this hypothesis on carunclectomized ewes was inconclusive (Owens et al., 1989). Ewes were allowed ad libitum access to feed and intakes were not reported. This presumably contributed to the highly variable rates of maternal glucose production in this study, and a confounding trend for these rates to be lower in carunclectomized than in control ewes.

Table 16.1. Effect of switching maternal nutrient intake at day 50 of gestation on placental development and fetal growth in monotocous adolescent ewes. Values are means for 11 or 12 ewes. Data from Wallace et al. (1999). Plane of nutrition1 Variable Placental weight (g) Placentome number Birth weight (kg)

MM 457a 103a 4.94a

HH

HM

MH

258b

381ac

312bc

79b 3.03b

96a 4.45a

109a 3.11b

PSE2 44 6.3 0.39

1M-, moderate intake, growth rate 57 g day⫺1, days 0–50; -M, moderate intake, days 50–100; H-, high intake, growth rate 280 g day⫺1, days 0–50; -H, high intake, days 50–100. All ewes were fed to maintain body condition from day 100 to term. 2 PSE, pooled standard error. Row means with different letters are significantly different (P < 0.05).

280

A.W. Bell and R.A. Ehrhardt

Placental transport and metabolism of nutrients Mechanisms of placental transport of glucose and amino acids Glucose Glucose, which accounts for about 60% of the net uptake of carbon by the gravid uterus in the well-nourished, late-pregnant ewe (Carver and Hay, 1995; Chung et al., 1998), is placentally transported from the maternal to the fetal circulations by carriermediated, facilitated diffusion (see Hay, 1995). The predominant glucose transporter protein isoforms in sheep placenta are GLUT-1 and GLUT-3 (Ehrhardt and Bell, 1997; Das et al., 1998), mRNA and protein abundance of which increase with gestational development, especially in GLUT-3 (Ehrhardt and Bell, 1997; Currie et al., 1997). This appears to account for much of the fivefold increase in glucose transport capacity of the sheep placenta in vivo between mid and late gestation (Molina et al., 1991). The relative importance of GLUT-3 may also be inferred from its immunolocalization at the apical surface of the trophoblastic cell layer that forms the fetomaternal tissue barrier in the ovine placenta (U.G. Das, R.A. Ehrhardt, W.W. Hay, Jr and S.U. Devaskar, unpublished observations). In contrast, GLUT-1 was localized on the basolateral surface of this cell layer, and in the plasma membrane of chorionic binucleate cells within the maternal caruncular epithelium. This differentially polarized localization of the two isoforms is consistent with the notion that GLUT-3, with a lower Km, efficiently transfers maternal glucose into trophoblastic cells surrounding the fetal vasculature. Exit of accumulated glucose from these cells into the umbilical microcirculation may then be mediated by GLUT-1, which has a higher Km and basolateral distribution. Negligible abundance of the insulin-responsive transporter protein, GLUT-4 (Ehrhardt and Bell, 1997), together with minimal expression of its mRNA in the sheep placenta (R.A. Ehrhardt, unpublished observations), is consistent with the lack of a direct effect of maternal or fetal insulinaemia on uteroplacental uptake and placental transport of glucose in vivo in the pregnant ewe (see Bell, 1993; Hay, 1995). Amino acids Most amino acids taken up by the sheep placenta are transported against a fetal– maternal concentration gradient, implying the use of energy-dependent, active transport processes (Carter et al., 1991). The specific protein transporters involved have yet to be identified and characterized in ruminant placentae. However, it is assumed that, as in human placental microvesicles, there are at least ten sodium-dependent and sodium-independent transporter systems which have varying levels of activity at different placental membrane surfaces (see Hay, 1998). Recent investigations suggest that rapid maternal–fetal transport of neutral amino acids may require not only sodium-dependent transport at the maternal surface of the placental epithelial surface, but affinity for highly reversible, sodium-independent transporters located at the fetal surface (Jozwik et al., 1998). This interpretation refutes the long-held assumption that passive diffusion is responsible for transporting neutral amino acids from placenta to fetus down an intracellular–extracellular concentration gradient created by sodium-dependent transporters (see Bell, 1993). Fractional exchange of all amino acids across the uterine circulation is small (£ 10%) in well-fed, late-pregnant ewes, implying that, under normal conditions,

Regulation of Macronutrient Partitioning

281

maternal supply is not a limiting factor for uterine uptake of amino acids (Chung et al., 1998). Negligible uterine uptake of glycine, which is most abundant in maternal blood, and glutamate suggest that the sodium-dependent transport systems for these acids, observed in other tissues, are not expressed in the ovine placenta.

Impact of placental metabolism on maternal–fetal nutrient transfer Glucose metabolism Glucose entry into the gravid uterus and its component tissues is determined by maternal arterial glucose concentration (Hay and Meznarich, 1988; Leury et al., 1990), while glucose transport to the fetus is determined by the transplacental (maternal–fetal) concentration gradient (Hay et al., 1984). In turn, the transplacental gradient is directly related to both placental and fetal glucose consumption, which are dependent on fetal arterial glucose concentration (Hay et al., 1990). Thus, as fetal glucose concentration changes relative to that of the mother, thereby changing the transplacental gradient, placental transfer of glucose to the fetus varies reciprocally with placental glucose consumption. In addition to its quantitative impact on placental transfer of glucose, placental glucose metabolism has a major qualitative influence on the pattern of carbohydrate metabolites delivered to the fetus. Rapid metabolism to lactate (~35%), fructose (~4%), and CO2 (~17%) accounted for about 56% of uteroplacental glucose consumption in late-pregnant ewes, and was directly related to placental glucose supply (Aldoretta et al., 1994). The fate of the remaining 44% of glucose metabolized by the placenta must include synthesis of alanine and other non-essential amino acids (Timmerman et al., 1998), directly or via lactate (Carter et al., 1995). Umbilical uptake and fetal oxidation of placentally derived lactate (Sparks et al., 1982; Hay et al., 1983a) and fructose (Meznarich et al., 1987) are estimated to contribute approximately 20% and 5%, respectively, to fetal CO2 production, in addition to the 30% contributed by the rapid oxidation of glucose (Hay et al., 1983a). Amino acid metabolism Placental metabolism substantially affects both the quantity and composition of amino acids delivered to the fetus. The significant net consumption by uteroplacental tissues of glutamate, serine, and the branched chain amino acids (Liechty et al., 1991; Chung et al., 1998) implies catabolism or transamination of these acids. An additional, small fraction of this net loss of amino acids will be in the form of secreted peptides. The ovine placenta has very little enzymatic capacity for urea synthesis but produces considerable amounts of ammonia, much of which is released into maternal and, to a lesser extent, fetal circulations (Holzman et al., 1977; Bell et al., 1989a). This is consistent with extensive placental deamination of branched chain amino acids to their respective keto acids, which are released into fetal and maternal bloodstreams (Smeaton et al., 1989; Loy et al., 1990), and with rapid rates of glutamate oxidation in the placenta (Moores et al., 1994). Transamination of branched chain amino acids accounts for some of the net glutamate acquisition by the placenta, the remainder of which is taken up from the umbilical circulation (Moores et al., 1994). That which is not quickly oxidized combines with ammonia to synthesize glutamine, which is then

282

A.W. Bell and R.A. Ehrhardt

released back into the umbilical bloodstream (Chung et al., 1998). Some of this glutamine is converted back to glutamate by the fetal liver, which produces most of the glutamate consumed by the placenta (Vaughn et al., 1995). This establishes a glutamate–glutamine shuttle which promotes placental oxidation of glutamate and fetal hepatic utilization of the amide group of glutamine. Similarly, the placenta almost quantitatively converts serine, mostly taken up from maternal blood, to glycine (Chung et al., 1998), reconciling the discrepancy between the negligible net uptake of glycine by the uterus and substantial net release of this amino acid into the umbilical circulation (see Hay, 1998). The complexity of interrelations among placental uptake, metabolism, and transport of amino acids was further illustrated by a recent study of alanine metabolism in ewes during late pregnancy (Timmerman et al., 1998). Application of tracer methodology showed that negligible net placental consumption or production of alanine masks an appreciable metabolism of maternal alanine entering the placenta which exchanges with endogenously produced alanine. Thus, most of the alanine delivered to the fetus is of placental origin, derived from placental protein turnover and transamination.

Factors affecting placental capacity for nutrient transport Placental size Placental capacity for glucose transport was substantially reduced, as were uteroplacental glucose consumption rate and fetal glycaemia, in carunclectomized (Owens et al., 1987a) and heat-treated ewes (Bell, 1987; Thureen et al., 1992). At least part of the absolute reduction in glucose transport capacity is presumed to be due to reduction in exchange surface area of the trophoblastic membrane, as shown in carunclectomized ewes (Robinson et al., 1995). In previously heat-treated ewes (Thureen et al., 1992), placental weight-specific glucose transport capacity was also reduced. This implies that chronic heat stress, which reduces average weight but not total number of placentomes, additionally reduces number and/or activity of specific glucose transport proteins at maternal and/or fetal exchange surfaces. In contrast, carunclectomy, which reduces placentome number but may stimulate a compensatory increase in average weight of individual placentomes, caused a modest increase in the placental weight-specific clearance of the non-metabolizable glucose analogue, 3-O-methyl glucose (Owens et al., 1987b). This implies that glucose transporter expression was preserved or increased in the remaining placentomes. Placental insufficiency in heat-treated ewes also extends to impaired capacity for amino acid transport, including major reductions in placental uptake and fetal transfer of leucine (Ross et al., 1996) and threonine (Anderson et al., 1997). The normally extensive placental catabolism of leucine was also greatly reduced (Ross et al., 1996). Maternal nutrition The placenta appears to play an active role in ameliorating the effects of maternal undernutrition on glucose partitioning to the conceptus. We recently found that restriction of the energy intake of ditocous ewes to 60% of predicted requirements for 2 weeks caused moderate maternal and fetal hypoglycaemia and a 26% reduction in the maternal–fetal glucose concentration gradient (Table 16.2). This was offset by a 50%

Regulation of Macronutrient Partitioning

283

Table 16.2. Maternal weight change, fetal weight, and indices of placental glucose transport at day 135 of pregnancy in ditocous ewes fed 100% or 60% of predicted energy requirements for the preceding 14 days (R.A. Ehrhardt and A.W. Bell, unpublished observations). Energy intake, % requirement a Variable ⌬ maternal weight (kg) Fetal weight (kg) Plasma glucose (mM) Maternal Fetal ⌬ maternal–fetal Placental 3MGe clearance (ml min⫺1 kg⫺1 placenta) CBf sites (pmol mg⫺1 protein) GLUT-3 protein (arbitrary unitsg)

60

PSEb

Pc

5.3 3.58

⫺2.7 3.46

1.0 0.16

< 0.001 NSd

3.72 0.57 3.15 117

2.84 0.49 2.33 176

0.09 0.03 0.03 7

< 0.001 < 0.05 < 0.001 < 0.001

105 1.00

126 1.19

3 0.04

< 0.01 < 0.05

100

a Values

are means for five ewes and ten conceptuses. pooled standard error. c P, significance of difference between means. d NS, not significant. e 3MG, 3-O-methyl glucose. f CB, cytochalasin B binding sites. g Expressed relative to control (100% energy intake). b PSE,

increase in maternal–fetal glucose transport capacity in vivo, at least partly explained by a 20% increase in placental glucose transporter abundance in vitro, assessed by binding of cytochalasin B and GLUT-3 protein abundance (Table 16.2). Possible adaptations in cellular distribution, in addition to total abundance, of placental GLUT-3 remain to be investigated. The lack of impairment of fetal growth attests to the success of adaptive responses to maintain placental glucose transfer (Table 16.2). During more severe maternal undernutrition or starvation for several days, the development of profound fetal hypoglycaemia helps to sustain the maternal–fetal gradient in glucose concentration by restricting the reverse transfer of glucose to the placenta, and reducing placental glucose consumption (Hay, 1995). More specific manipulation of maternal and fetal glycaemia by prolonged maternal infusion with insulin has shown that the decline in fetal glucose concentration is less than that of the mother. This tends to decrease the maternal–fetal glucose concentration gradient, protecting placental glucose consumption at the expense of the fetus. In response, fetal glucose needs are diminished by a reduction in fetal growth rate (Carver and Hay, 1995), associated with decreased placental transfer and fetal accretion of leucine, and increased release of leucine from proteolysis in fetal tissues (Carver et al., 1997).

284

A.W. Bell and R.A. Ehrhardt

Metabolic adaptations in maternal tissues Pregnancy-specific effects on glucose and amino acid metabolism Maternal strategies for accommodating the substantial requirements for glucose and amino acids of the conceptus include adaptations in carbohydrate and protein metabolism, briefly discussed below. They also involve changes in lipid metabolism which result in increased mobilization and maternal metabolism of NEFAs (see Bell and Bauman, 1994). Although maternal NEFAs are not readily used by the conceptus, their increased utilization in maternal tissues serves to spare maternal glucose and, perhaps, amino acids to meet requirements of the conceptus. Glucose metabolism Hepatic gluconeogenesis increases in ewes during late pregnancy even when feed intake is not increased above non-pregnant levels (Freetly and Ferrell, 1998), consistent with earlier observations of pregnancy-specific effects on whole-body glucose kinetics (see Bell, 1993). This was concomitant with increased hepatic uptake of lactate (Freetly and Ferrell, 1998), apparently derived from uteroplacental metabolism (Meschia et al., 1980; Faichney et al., 1981) and increased glycolysis in maternal peripheral tissues (Hough et al., 1985). Part of the moderate net mobilization of amino acids from carcass tissues of late-pregnant ewes fed to predicted nutrient requirements (Fig. 16.1; McNeill et al., 1997) may also be used to sustain increased hepatic gluconeogenesis as term approaches. Glucose uptake by maternal peripheral tissues such as hindlimb muscle and adipose tissue tends to decline during late pregnancy, although the evidence for some of these responses in ‘well-fed’ animals may have been confounded by variations in voluntary feed intake (see Bell and Bauman, 1997). Amino acid metabolism Effects of pregnancy on the quantitative metabolism of amino acids have not been systematically studied in ruminants. However, in agreement with data from rats (Ling et al., 1987), the fractional rate of hepatic protein synthesis increases 45% during late pregnancy in dairy cows (Bell, 1995). This is consistent with the moderate increase in hepatic protein accretion (Campbell and Fell, 1970), and an apparent decrease in hepatic amino acid catabolism (Freetly and Ferrell, 1998) in late-pregnant ewes. These changes in hepatic metabolism occurred despite a decreased (dairy cows) or unchanged (ewes) protein intake and hepatic uptake of amino acids, implying endogenous regulation of hepatic disposal of amino acids as in late-pregnant rats (Casado et al., 1987).

Modulation by maternal nutrition Energy Most of the relatively modest metabolic adaptations that occur in maternal tissues in well-fed pregnant ruminants are readily exacerbated or modified by maternal energy intake, especially during late pregnancy. Thus, the tendency for maternal insulinresponsive peripheral tissues to reduce glucose utilization during late pregnancy is

Regulation of Macronutrient Partitioning

285

exaggerated by moderate underfeeding, consistent with reductions in abundance of GLUT-4 protein in skeletal muscle and perirenal adipose tissue of ewes fed at 60% of predicted requirements for 2 weeks in late pregnancy (Fig. 16.2; Ehrhardt et al., 1998). On the other hand, the decline in maternal glucose production in moderately underfed ewes was not as great as predicted by their reduction in energy intake (Petterson et al., 1993). This, together with the fact that glucose production in energyrestricted pregnant ewes was 40% greater than that in non-pregnant ewes consuming 13% more energy (Petterson et al., 1993), implies that hepatic gluconeogenesis in underfed pregnant ewes was supported by much greater reliance on endogenous glucogenic substrates such as amino acids, glycerol and lactate. Mobilization of amino acids from skeletal muscle to meet this need is consistent with the shrinkage of muscle fibres in dairy cows underfed during late pregnancy (Reid et al., 1980). Protein As discussed earlier, protein deprivation of ditocous ewes during late pregnancy increased the net loss of protein from carcass tissues and abolished the normal, pregnancy-related increase in protein accretion of visceral organs. Conversely, feeding protein at levels above predicted requirements (160 g kg⫺1 versus 120 g kg⫺1 dry matter), caused an appreciable reversal of the net flux of nitrogen from maternal carcass tissues and allowed significant net accretion of tissue protein during late pregnancy (Fig. 16.1; McNeill et al., 1997).

Fig. 16.2. Relative abundance of GLUT-4 protein in perirenal (PR) adipose tissue and semitendinosus (ST) muscle at day 135 of pregnancy in ditocous ewes that had been fed at 100% or 60% of predicted energy requirements for the preceding 14 days. Histograms are means for five ewes, expressed relative to control values (100% energy intake) for each tissue. Pooled standard errors were 0.07 units for PR adipose, and 0.06 units for ST muscle. Within tissues, means with different letters are significantly different (P < 0.05). From the data of Ehrhardt et al. (1998).

286

A.W. Bell and R.A. Ehrhardt

Homeorhetic regulation General concept The concept of homeorhesis as applied to regulation of nutrient partitioning (Bauman and Currie, 1980) is discussed elsewhere in this volume (Chapter 18). Key postulates of this concept include its simultaneous influence on multiple tissues and functional systems, implying extracellular mediation, and its operation through altered tissue responses to homeostatic effectors such as insulin, at various levels of extracellular and intracellular signalling. Altered tissue responses to insulin Application of the euglycaemic, hyperinsulinaemic clamp technique has demonstrated that in sheep, as in humans and laboratory animals, late pregnancy is associated with moderate insulin resistance, and that this condition is exacerbated by undernutrition. This was manifested as diminished sensitivity to insulin of several variables of wholebody glucose utilization (Petterson et al., 1993) and decreased insulin responsiveness of lipolysis and NEFA mobilization (Petterson et al., 1994). Tissue specificity of these whole-body phenomena was not assessed in these studies, but it is likely that adaptations in the major insulin-responsive peripheral tissues, skeletal muscle and adipose tissue, were mostly responsible (see Bell and Bauman, 1997). As pregnancy advances, ovine adipose tissue in vitro becomes refractory to the stimulation of lipogenesis by insulin (Vernon et al., 1985; Guesnet et al., 1991). Although glucose is not an important carbon precursor for lipogenesis in ruminants, it is required for synthesis of glyceride glycerol, and for part of the NADPH necessary for de novo fatty acid synthesis (see Bauman and Davis, 1975). This is consistent with our recent observation of progressive development of insulin resistance in terms of wholebody glucose disposal during pregnancy in fat-tailed Karakul ewes (Slepetis et al., 1999). The role of skeletal muscle in the evolution of insulin resistance during pregnancy has not been addressed in ruminants. However, it seems likely that the ability of insulin to promote glucose uptake by muscle is diminished in late pregnancy as in early lactation (Vernon et al., 1990) because of the similar degrees of whole-body insulin resistance observed in ewes in these two states (Slepetis et al., 1999). The reduction in muscle abundance of the insulin-responsive GLUT-4 protein in underfed versus wellfed pregnant ewes (Fig. 16.2; Ehrhardt et al., 1998) provides a mechanistic clue. Possible homeorhetic agents Various hormones, including progesterone, oestradiol and placental lactogen (PL), may act as homeorhetic regulators of observed changes in tissue responses to insulin and associated metabolic adaptations to the state of pregnancy in ruminants (Bell and Bauman, 1994, 1997). The argument for PL is especially hard to dismiss, despite a continuing lack of direct experimental evidence. This uniquely placental peptide crossreacts with both growth hormone (GH) and prolactin receptors in ruminant tissues (Byatt et al., 1992), and its specific binding in adipose tissue increases with advancing pregnancy in sheep (N’Guema et al., 1986). However, it remains unclear whether highaffinity binding of PL in adipose tissue or liver is mediated through structurally unique receptors (Anthony et al., 1995). Cross-reactivity with the GH receptor would be con-

Regulation of Macronutrient Partitioning

287

sistent with the development of insulin resistance in adipose tissue during late pregnancy because GH is a potent homeorhetic effector of this response in ruminant adipose tissue (see Bell and Bauman, 1997; Chapter 18). Other indirect evidence for a homeorhetic role for PL includes enhanced placental gene expression and secretion of ovine placental lactogen in moderately undernourished, late-pregnant ewes, coincident with decreased expression of GLUT-4 in maternal insulin-responsive tissues (Fig. 16.2; Ehrhardt et al., 1998) and exaggeration of indices of whole-body insulin resistance discussed earlier in this section. It is also possible that leptin, the ob gene product expressed almost exclusively in adipose tissue (see Houseknecht et al., 1998), plays a role in the homeorhetic coordination of conceptus nutrient demands and metabolic adaptations in maternal tissues of pregnant ruminants. In addition to its commonly postulated role as a signal of peripheral energy status to the central nervous system, leptin may have pleiotropic effects on peripheral tissues, including mediation of insulin resistance in adipocytes (Muller et al., 1997). Coincident with the evolution of insulin resistance, we have recently observed a consistent, threefold increase in expression of leptin mRNA in tail adipose tissue from fat-tailed Karakul ewes during mid (50–60 days) and late (125–135 days) pregnancy compared with levels in the same animals when non-pregnant (Fig. 16.3). A similar, pregnancy-induced increase in leptin mRNA expression in white adipose tissue was associated with a marked increase in serum concentrations of leptin in mice (Tomimatsu et al., 1997). Leptin may also act on the placenta, which, in sheep, strongly expresses the leptin receptor splice variant OB-Rb (Ehrhardt et al., 1999) that is considered essential for intracellular signal transduction after binding leptin. This raises the possibility that, in

Fig. 16.3. Relative abundance of leptin mRNA in tail adipose tissue of Karakul ewes (n = 8) sequentially sampled while non-pregnant and non-lactating (NP), in mid pregnancy (MP, day 50–60), in late pregnancy (LP, day 125–135), and in early lactation (L, day 15–22 post-partum). Pooled standard error was 0.46 units. Means with different letters are significantly different (P < 0.05). From the unpublished data of R.A. Ehrhardt, R.M. Slepetis, Y.R. Boisclair and A.W. Bell.

288

A.W. Bell and R.A. Ehrhardt

addition to its putative influences on adipose and other maternal tissues, leptin may directly or indirectly mediate observed adaptations in placental capacity for glucose transport during advancing gestation and moderate maternal undernutrition.

Conclusions We propose that the placenta plays a pivotal, multi-faceted role in the regulation of nutrient partitioning between maternal tissues and the conceptus, especially during late pregnancy when most of fetal growth occurs. First, it is a highly regulated nutritional and excretory conduit that provides a means of coordinating fetal nutrient demands with maternal capacity to supply these nutrients, especially glucose and amino acids. Under optimal conditions, this will result in appropriate constraint of late-gestation fetal growth such that birth weight will be sufficient to permit a good chance of neonatal survival, but insufficient to cause dystocia or an inappropriate depletion of maternal nutrient reserves prior to the onset of lactation. The mechanistic bases for this general role of the placenta are beginning to be understood, although much of the details about amino acid transport and its regulation have yet to be elucidated. Second, through its capacity to synthesize and secrete a plethora of bioactive molecules, including proteins, steroids, and eicosanoids, into maternal and fetal circulations, the placenta offers unique opportunities for direct and indirect communication between the conceptus and its maternal host. Although lacking direct evidence, the hypothesis that PL and/or other placental hormones are responsible for effecting insulin resistance and other metabolic adaptations in maternal tissues remains attractive. We also speculate that communication of maternal energy status to the placenta may be mediated through the leptin system, adding a further layer of complexity to homeorhetic coordination of maternal nutrition and energy stores with nutrient requirements of the conceptus.

References Aldoretta, P.W., Anderson, S. and Hay, W.W., Jr (1994) Ovine uteroplacental glucose and oxygen metabolism in relation to chronic changes in maternal and fetal glucose concentrations. Placenta 15, 753–764. Anderson, A.H., Fennessey, P.V., Meschia, G., Wilkening, R.B. and Battaglia, F.C. (1997) Placental transport of threonine and its utilization in the normal and growth-restricted fetus. American Journal of Physiology 272, E892–E900. Anthony, R.V., Pratt, S.L., Liang, R. and Holland, M.D. (1995) Placental-fetal hormonal interactions: impact on fetal growth. Journal of Animal Science 73, 1861–1871. Battaglia, F.C. and Meschia, G. (1988) Fetal nutrition. Annual Review of Nutrition 8, 43–61. Bauman, D.E. and Currie, W.B. (1980) Partitioning of nutrients during pregnancy and lactation: a review of mechanisms involving homeostasis and homeorhesis. Journal of Dairy Science 63, 1514–1529. Bauman, D.E. and Davis, C.L. (1975) Regulation of lipid metabolism. In: McDonald, I.W. and Warner, A.C.I. (eds) Digestion and Metabolism in the Ruminant. The University of New England Publishing Unit, Armidale, pp. 496–509. Bell, A.W. (1987) Consequences of severe heat stress for fetal development. In: Hales, J.R.S. and

Regulation of Macronutrient Partitioning

289

Richards, D.A.B. (eds) Heat Stress: Physical Exertion and Environment. Elsevier, Amsterdam, pp. 313–333. Bell, A.W. (1993) Pregnancy and fetal metabolism. In: Forbes, J.M. and France, J. (eds) Quantitative Aspects of Ruminant Digestion and Metabolism. CAB International, Wallingford, pp. 405–431. Bell, A.W. (1995) Regulation of organic nutrient metabolism during transition from late pregnancy to early lactation. Journal of Animal Science 73, 2804–2819. Bell, A.W and Bauman, D.E. (1994) Animal models for the study of adipose regulation in pregnancy and lactation. In: Allen, L., King, J. and Lonnerdal, B. (eds) Nutrient Regulation during Pregnancy, Lactation, and Infant Growth. Plenum Press, New York, pp. 71–84. Bell, A.W. and Bauman, D.E. (1997) Adaptations of glucose metabolism during pregnancy and lactation. Journal of Mammary Gland Biology and Neoplasia 2, 265–278. Bell, A.W. and Ehrhardt, R.A. (1998) Placental regulation of nutrient partitioning during pregnancy. In: Hansel, W., Bray, G.A. and Ryan, D.H. (eds) Nutrition and Reproduction. Louisiana State University Press, Baton Rouge, pp. 229–254. Bell, A.W., Kennaugh, J.M., Battaglia, F.C. and Meschia, G. (1989a) Uptake of amino acids and ammonia at mid-gestation by the fetal lamb. Quarterly Journal of Experimental Physiology 74, 635–643. Bell, A.W., McBride, B.W., Slepetis, R., Early, R.J. and Currie, W.B. (1989b) Chronic heat stress and prenatal development in sheep: 1. Conceptus growth and maternal plasma hormones and metabolites. Journal of Animal Science 67, 3289–3299. Bell, A.W., Slepetis, R. and Ehrhardt, R.A. (1995) Growth and accretion of energy and protein in the gravid uterus during late pregnancy in Holstein cows. Journal of Dairy Science 78, 1954–1961. Byatt, J.C., Warren, W.C., Eppard, P.J., Staten, N.R., Krivi, G.G. and Collier, R.J. (1992) Ruminant placental lactogens: structure and biology. Journal of Animal Science 70, 2911–2923. Campbell, R.M. and Fell, B.F. (1970) Observations on hypertrophy of the liver in breeding ewes. Research in Veterinary Science 11, 540–547. Carter, B.S., Moores, R.R. and Battaglia, F.C. (1991) Placental transport and fetal and placental metabolism of amino acids. Journal of Nutritional Biochemistry 2, 4–13. Carter, B.S., Moores, R.R., Jr, Teng, C., Meschia, G. and Battaglia, F.C. (1995) Main routes of plasma lactate carbon disposal in the midgestation fetal lamb. Biology of the Neonate 67, 295–300. Carver, T.D. and Hay, W.W., Jr (1995) Uteroplacental carbon substrate metabolism and O2 consumption after long-term hypoglycemia in pregnant sheep. American Journal of Physiology 269, E299–E308. Carver, T.D., Quick, A.A., Teng, C.C., Pike A.W., Fennessey, P.V. and Hay, W.W., Jr (1997) Leucine metabolism in chronically hypoglycemic hypoinsulinemic growth-restricted fetal sheep. American Journal of Physiology 272, E107–E117. Casado, J., Remesar, X. and Pastor-Anglada, M. (1987) Hepatic uptake of amino acids in latepregnant rats. Biochemical Journal 248, 117–122. Chung, M., Teng, C., Timmerman, M., Meschia, G. and Battaglia, F.C. (1998) Production and utilization of amino acids by ovine placenta in vivo. American Journal of Physiology 274, E13–E22. Clarke, L., Heasman, L., Juniper, D.T. and Symonds, M.E. (1998) Maternal nutrition in earlymid gestation and placental size in sheep. British Journal of Nutrition 79, 359–364. Comline, R.S. and Silver, M. (1976) Some aspects of foetal and utero-placental metabolism in cows with indwelling umbilical and uterine vascular catheters. Journal of Physiology 260, 571–586. Currie, M.J., Bassett, N.S. and Gluckman, P.D. (1997) Ovine glucose transporter-1 and -3: cDNA partial sequences and developmental gene expression in the placenta. Placenta 18, 393–401.

290

A.W. Bell and R.A. Ehrhardt Das, U.G., Sadiq, H.F., Soares, M.J., Hay, W.W., Jr and Devaskar, S.U. (1998) Time-dependent physiological regulation of rodent and ovine placental glucose transporter (GLUT-1) protein. American Journal of Physiology 274, R339–R347. Ehrhardt, R.A. and Bell, A.W. (1995) Growth and metabolism of the ovine placenta during midgestation. Placenta 16, 727–741. Ehrhardt, R.A. and Bell, A.W. (1997) Developmental increases in glucose transporter concentration in the sheep placenta. American Journal of Physiology 273, R1132–R1141. Ehrhardt, R.A., Slepetis, R.M. and Bell, A.W. (1998) Moderate undernutrition alters glucose transporter levels in maternal insulin-responsive tissues and in the placenta. Journal of Animal Science 76 (Suppl. 1), 130. Ehrhardt, R.A., Bell, A.W. and Bosclair, Y.R. (1999) Analysis of leptin and leptin RNA expression in the sheep placenta. Journal of Animal Science 77 (suppl. 1), 159. Faichney, G.J. and White, G.A. (1987) Effects of maternal nutritional status on fetal and placental growth and on fetal urea synthesis in sheep. Australian Journal of Biological Sciences 40, 365–377. Faichney, G.J., Barker, P.J., Setchell, B.P. and Lindsay, D.B. (1981) The utilization of lactic acid by sheep in late pregnancy. Quarterly Journal of Experimental Physiology 66, 195–201. Falconer, J., Owens, J.A., Allotta, E. and Robinson, J.S. (1985) Effect of restriction of placental growth on the concentrations of insulin, glucose and placental lactogen in the plasma of sheep. Journal of Endocrinology 106, 7–11. Ferrell, C.L. (1991) Maternal and fetal influences on uterine and conceptus development in the cow: II. Blood flow and nutrient flux. Journal of Animal Science 69, 1954–1965. Freetly, H.C. and Ferrell, C.L. (1998) Net flux of glucose, lactate, volatile fatty acids, and nitrogen metabolites across the portal-drained viscera and liver of pregnant ewes. Journal of Animal Science 76, 3133–3145. Guesnet, P.M., Massoud, M.J. and Demarne, Y. (1991) Regulation of adipose tissue metabolism during pregnancy and lactation in the ewe: the role of insulin. Journal of Animal Science 69, 2057–2065. Hammond, J. (1947) Animal breeding in relation to nutrition and environmental conditions. Biological Reviews 22, 195–213. Hay, W.W., Jr (1995) Regulation of placental metabolism by glucose supply. Reproduction, Fertility and Development 7, 365–375. Hay, W.W., Jr (1996) Nutrition and development of the fetus: carbohydrate and lipid metabolism. In: Walker, W.A. and Watkins, J.B. (eds) Nutrition in Pediatrics. B.C. Decker, Hamilton, pp. 364–378. Hay, W.W., Jr (1998) Fetal requirements and placental transfer of nitrogenous compounds. In: Polin, R.A. and Fox, W.W. (eds) Fetal and Neonatal Physiology, 2nd edn. Saunders, Philadelphia, pp. 619–634. Hay, W.W., Jr and Meznarich, H.K. (1988) Effect of maternal glucose concentration on uteroplacental glucose consumption and transfer in pregnant sheep. Proceedings of the Society for Experimental Biology and Medicine 190, 63–69. Hay, W.W., Jr, Myers, S.A., Sparks, J.W., Wilkening, R.B., Meschia, G. and Battaglia, F.C. (1983a) Glucose and lactate oxidation in the fetal lamb. Proceedings of the Society for Experimental Biology and Medicine 173, 553–563. Hay, W.W., Jr, Sparks, J.W., Wilkening, R.B., Battaglia, F.C. and Meschia, G. (1983b) Partition of maternal glucose production between conceptus and maternal tissues in sheep. American Journal of Physiology 245, E347–E350. Hay, W.W., Jr, Sparks, J.W., Wilkening, R.B., Battaglia, F.C. and Meschia, G. (1984) Fetal glucose uptake and utilization as functions of maternal glucose concentration. American Journal of Physiology 246, E237–E242. Hay, W.W., Jr, Molina, R.D., DiGiacamo, J.E. and Meschia, G. (1990) Model of placental glucose consumption and transfer. American Journal of Physiology 258, R569–R577.

Regulation of Macronutrient Partitioning

291

Holzman, I.R., Lemons, J.A., Meschia, G. and Battaglia, F.C. (1977) Ammonia production by the pregnant uterus. Proceedings of the Society for Experimental Biology and Medicine 156, 27–30. Hough, G.M., McDowell, G.H., Annison, E.F. and Williams, A.J. (1985) Glucose metabolism in hindlimb muscle of pregnant and lactating ewes. Proceedings of the Nutrition Society of Australia 10, 97. Houseknecht, K.L., Baile, C.A., Matteri, R.L. and Spurlock, M.E. (1998) The biology of leptin: a review. Journal of Animal Science 76, 1405–1420. Jozwik, M., Teng, C., Timmerman, M., Chung, M., Meschia, G. and Battaglia, F.C. (1998) Uptake and transport by the ovine placenta of neutral nonmetabolizable amino acids with different transport system affinities. Placenta 19, 531–538. Lemons, J.A. and Schreiner, R.L. (1983) Amino acid metabolism in the ovine fetus. American Journal of Physiology 244, E459–E466. Leury, B.J., Bird, A.R., Chandler, K.D. and Bell, A.W. (1990) Glucose partitioning in the pregnant ewe: effects of undernutrition and exercise. British Journal of Nutrition 64, 449–462. Liechty, E.A., Kelley, J. and Lemons, J.A. (1991) Effect of fasting on uteroplacental amino acid metabolism in the pregnant sheep. Biology of the Neonate 60, 207–214. Ling, P.R., Bistrian, B.R., Blackburn, G.L. and Istfan, N. (1987) Effect of fetal growth on maternal metabolism in postabsorptive rat. American Journal of Physiology 252, E380–E390. Loy, G.L., Quick, A.N., Jr, Hay, W.W., Jr, Meschia, G., Battaglia, F.C. and Fennessey, P.V. (1990) Fetoplacental deamination and decarboxylation of leucine. American Journal of Physiology 259, E492–E497. McCrabb, G.J., Egan, A.R. and Hosking, B.J. (1992) Maternal undernutrition during midpregnancy in sheep: variable effects on placental growth. Journal of Agricultural Science, Cambridge 118, 127–132. McNeill, D.M., Slepetis, R., Ehrhardt, R.A., Smith, D.M. and Bell, A.W. (1997) Protein requirements of sheep in late pregnancy: partitioning of nitrogen between gravid uterus and maternal tissues. Journal of Animal Science 75, 809–816. McNeill, D.M., Kelly, R.W. and Williams, I.H. (1998) Partition of nutrients in moderately fat ewes compared with lean ewes given ad libitum access to feed in late pregnancy. Australian Journal of Agricultural Research 49, 575–580. McNeill, D.M., Kelly, R.W. and Williams, I.H. (1999) Maternal fatness influences fetal size in ewes underfed in late pregnancy. Australian Journal of Agricultural Research 50, 1171–1177. Meschia, G., Battaglia, F.C., Hay, W.W., Jr and Sparks, J.W. (1980) Utilization of substrates by the ovine placenta in vivo. Federation Proceedings 39, 245–249. Meznarich, H.K., Hay, W.W., Jr, Sparks, J.W., Meschia, G. and Battaglia, F.C. (1987) Fructose disposal and oxidation rates in the ovine fetus. Quarterly Journal of Experimental Physiology 72, 617–625. Molina, R.D., Meschia, G., Battaglia, F.C. and Hay, W.W., Jr (1991) Gestational maturation of placental glucose transfer capacity in sheep. American Journal of Physiology 261, R697–R704. Moores, R.R., Jr, Vaughn, P.R., Battaglia, F.C., Fennessey, P.V., Wilkening, R.B. and Meschia, G. (1994) Glutamate metabolism in the fetus and placenta of late-gestation sheep. American Journal of Physiology 267, R89–R96. Muller, G., Ertl, J., Gerl, M. and Preibisch, G. (1997) Leptin impairs metabolic actions of insulin in isolated rat adipocytes. Journal of Biological Chemistry 272, 10585–10593. N’Guema, M., Delouis, C., Kelly, P.A. and Djiane, J. (1986) Evolution of prolactin and placental lactogen receptors in ewes during pregnancy and lactation. Endocrinology 118, 695–700. Oddy, V.H., Gooden, J.M., Hough, G.M., Teleni, E. and Annison, E.F. (1985) Partitioning of nutrients in Merino ewes. II Glucose utilization by skeletal muscle, the pregnant uterus and the lactating mammary gland in relation to whole body glucose utilization. Australian Journal of Biological Sciences 38, 95–108.

292

A.W. Bell and R.A. Ehrhardt Owens, J.A., Falconer, J. and Robinson, J.S. (1987a) Effect of restriction of placental growth on fetal and utero-placental metabolism. Journal of Developmental Physiology 9, 225–238. Owens, J.A., Falconer, J. and Robinson, J.S. (1987b) Restriction of placental size in sheep enhances efficiency of placental transfer of antipyrine, 3-O-methyl-D-glucose but not urea. Journal of Developmental Physiology 9, 457–464. Owens, J.A., Falconer, J. and Robinson, J.S. (1989) Glucose metabolism in pregnant sheep when placental growth is restricted. American Journal of Physiology 257, R350–R357. Petterson, J.A., Dunshea, F.R., Ehrhardt, R.A. and Bell, A.W. (1993) Pregnancy and undernutrition alter glucose metabolic responses to insulin in sheep. Journal of Nutrition 123, 1286–1295. Petterson, J.A., Slepetis, R., Ehrhardt, R.A., Dunshea, F.R. and Bell, A.W. (1994) Pregnancy but not moderate undernutrition attenuates insulin suppression of fat mobilization in sheep. Journal of Nutrition 124, 2431–2436. Prior, R.L. and Christenson, R.K. (1978) Insulin and glucose effects on glucose metabolism in pregnant and nonpregnant ewes. Journal of Animal Science 46, 201–210. Reid, I.M., Roberts, C.J. and Baird, G.D. (1980) The effects of underfeeding during pregnancy and lactation on structure and chemistry of bovine liver and muscle. Journal of Agricultural Science, Cambridge 94, 239–245. Reynolds, L.P., Ferrell, C.L., Robertson, D.A. and Ford, S.P. (1986) Metabolism of the gravid uterus, foetus and utero-placenta at several stages of gestation in cows. Journal of Agricultural Science, Cambridge 106, 437–444. Robinson, J., Chidzanja, S., Kind, K., Lok, F., Owens, P. and Owens, J. (1995) Placental control of fetal growth. Reproduction, Fertility and Development 7, 333–344. Ross, J.C., Fennessey, P.V., Wilkening, R.B., Battaglia, F.C. and Meschia, G. (1996) Placental transport and fetal utilization of leucine in a model of fetal growth retardation. American Journal of Physiology 270, E491–E503. Slepetis, R.M., Ehrhardt, R.A., Block, S.S., Boisclair, Y.R. and Bell, A.W. (1999) Alterations in in vivo whole-body insulin resistance according to physiological state in Karakul ewes. Journal of Animal Science 77 (Suppl. 1), 163. Smeaton, T.C., Owens, J.A., Kind, K.L. and Robinson, J.S. (1989) The placenta releases branched-chain amino acids into the umbilical and uterine circulations in the pregnant sheep. Journal of Developmental Physiology 12, 95–99. Sparks, J.W., Hay, W.W., Jr, Bonds, D., Meschia, G. and Battaglia, F.C. (1982) Simultaneous measurements of lactate turnover rate and umbilical lactate uptake in the fetal lamb. Journal of Clinical Investigation 70, 179–192. Thureen, P., Trembler, K.A., Meschia, G., Makowski, E.L. and Wilkening, R.B. (1992) Placental glucose transport in heat-induced fetal growth retardation. American Journal of Physiology 263, R578–R585. Timmerman, M., Chung, M., Wilkening, R.B., Fennessey, P.V., Battaglia, F.C. and Meschia, G. (1998) Relationship of fetal alanine uptake and placental alanine metabolism to maternal alanine concentration. American Journal of Physiology 275, E942–E950. Tomimatsu, T., Yamaguchi, M., Murakami, T., Ogura, K., Sakata, M., Mitsuda, N., Kanzaki, T., Kurachi, H., Irahara, M., Miyake, A., Shima, K., Aono, T. and Murata, Y. (1997) Increase of mouse leptin production by adipose tissue after midpregnancy: gestational profile of serum leptin concentration. Biochemical and Biophysical Research Communications 240, 213–215. van der Walt, J.G., Baird, G.D. and Bergman, E.N. (1983) Tissue glucose and lactate metabolism and interconversions in pregnant and lactating sheep. British Journal of Nutrition 50, 267–280. Vaughn, P.R., Lobo, C., Battaglia, F.C., Fennessey, P.V., Wilkening, R.B. and Meschia, G. (1995)

Regulation of Macronutrient Partitioning

293

Glutamine-glutamate exchange between placenta and fetal liver. American Journal of Physiology 268, E705–E711. Vernon, R.G., Clegg, R.A. and Flint, D.J. (1985) Adaptation of adipose tissue metabolism and number of insulin receptors in pregnant sheep. Comparative Biochemistry and Physiology 81B, 909–913. Vernon, R.G., Faulkner, A., Hay, W.W., Jr, Calvert, D.T. and Flint, D.J. (1990) Insulin resistance of hind-limb tissues in vivo in lactating sheep. Biochemical Journal 270, 783–786. Wallace, J.M., Aitken, R.P. and Cheyne, M.A. (1996) Nutrient partitioning and fetal growth in rapidly growing adolescent ewes. Journal of Reproduction and Fertility 107, 183–190. Wallace, J.M., Bourke, D.A., Aitken, R.P. and Cruickshank, M.A. (1999) Switching maternal dietary intake at the end of the first trimester has profound effects on placental development and fetal growth in adolescent ewes carrying singleton fetuses. Biology of Reproduction 61, 101–110.

17

The Thermal Physiology of the Ruminant Fetus

H. LABURN, A. FAURIE AND D. MITCHELL Department of Physiology and Brain Function Research Unit, University of the Witwatersrand, Parktown, Johannesburg, South Africa

Introduction We have come to know much about the thermal physiology of the ruminant fetus not because the thermal physiology of ruminants is peculiar, but because so much of the research on the physiology of fetal mammals has been carried out on sheep. That includes our own research. The experimental evidence, however, leads us to believe that much of what is known about the thermal physiology of the sheep fetus also applies to other ruminant species, and some non-ruminant species, including our own. It would be a reasonable expectation that the thermal relationship the fetal lamb has with the pregnant ewe is representative of the fetal–maternal relationship in other ruminants, where fetuses typically number one or two per pregnancy, and where total fetal mass constitutes approximately 10% maternal body mass at term.

Fetal body temperature homeostasis Fetal heat production and heat loss Fetal growth and development generates considerable metabolic heat. Oxygen consumption per kg of fetal body mass increases progressively during gestation with a peak prior to term (Bell et al., 1987), but even at term, the metabolic rate (about 3 W kg⫺1; Gilbert et al., 1985) of the sheep fetus is approximately twice that per kg body mass of the pregnant ewe or adult non-pregnant sheep (Abrams et al., 1970). Most fetal metabolic heat derives from the developing ‘vital organs’, including the brain, which although constituting only about 6% of body mass, contribute some 40% of total metabolic heat in late gestation (Bell et al., 1987). To avoid potentially dangerous heat storage, the fetus requires an efficient route for heat loss to dissipate the heat produced, and, as for many other metabolic by-products, it is the placenta which provides that route. Based on measurement of fetal lamb and pregnant ewe body temperatures, © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

295

296

H. Laburn et al.

and model predictions of the feto–maternal thermal relationship, Gilbert et al. (1985) estimated that approximately 85% of fetal heat is lost via the utero-placental circulation (Fig. 17.1). That heat is delivered to the placenta by the umbilical vessels, and leaves the placenta in the uterine circulation, so increases or decreases in umbilical and/or uterine blood flow will lead to increased or decreased heat transfer respectively, between the fetus and the mother animal (Schröder et al., 1988). The remainder of the heat produced is dissipated via convection and conduction through the amniotic fluid and subsequently the uterine wall. Ultimately, the heat is lost to the environment, via typical heat loss strategies employed by the pregnant animal, such as vasodilation in peripheral blood vessels, and panting.

Fetal body temperature The equilibrium between fetal heat production and heat loss via the placenta or uterine wall occurs when fetal body temperature is approximately 0.5°C higher than that of its mother’s body core; a situation which prevails in all species of mammal tested to date, including rabbits (Hart and Faber, 1965), baboons (Morishima et al., 1975), dogs (Assali and Westin, 1962), sheep (Gunn and Gluckman, 1983; Laburn et al., 1992)

Fig. 17.1. Diagrammatic representation of the routes of heat loss from the fetal lamb.

The Thermal Physiology of the Ruminant Fetus

297

and humans (Wood and Beard, 1964; Adamsons and Towell, 1965). Our measurements reveal typical values for body temperatures of the sheep fetus to be 39.6°C, and of the pregnant ewe to be 39.2°C. We recently have measured these temperatures in pregnant nanny goats and their fetuses, and have found almost identical values to those in sheep for fetal and maternal body temperatures in this ruminant species too. Why the equilibrium feto-maternal temperature difference is the same in species of different mass remains unknown.

The feto-maternal thermal relationship Body temperatures in laboratory and field conditions Not only is the feto-maternal (F-M) thermal gradient remarkably similar in different species, but it also appears to be remarkably constant, at least in sheep, in the face of potential thermal perturbations during the latter stages of gestation. Figure 17.2 shows measurements made using temperature-sensitive radiotelemeters, which had been implanted into the abdomens of a group of pregnant ewes and into the abdomens of their fetuses, over the last five weeks of gestation, when the animals were kept in thermally stable indoor conditions. On a day-to-day basis, the feto-maternal temperature relationship is very stable, despite large gains in fetal body mass, and metabolic heat produced (Bell et al., 1987). Moreover, Fig. 17.3 shows that, despite circadian fluctuations in the ewe’s body temperature, the F-M temperature difference does not show significant deviations on an hour-to-hour basis. In their natural environments or in field conditions, pregnant animals are subject to greater fluctuations in ambient conditions than are laboratory-housed animals, and one may ask whether the F-M thermal relationship is as constant then. Using temperature-sensitive data loggers, rather than radiotelemeters, implanted chronically into the fetal sheep and pregnant ewes, we compared the F-M relationship in laboratory conditions with that in field conditions. Figure 17.4 shows typical data obtained from one pregnant ewe. Ewe body core temperature varied much more in the field, with day- and night-time environmental temperature changes (Fig. 17.4b). Fetal body temperature changed too, but not nearly to the same extent as occurred in the mother animal. These data are the first recordings of which we are aware, of continuous body temperature measurements in a fetus under field conditions. Our measurements were made while the animals were sedentary and exposed to mild winter conditions – day-time dry-bulb temperature varied between 10° and 20°C, with relative humidity approximately 40%. Night-time temperatures were generally in the range 0–3°C. Measurements of fetal body temperature need to be made in animals living in harsher environments, to establish the robustness of the F-M thermal relationship, especially as many ruminants are bred and/or housed or live naturally in conditions considered to be extreme (Alexander and Williams, 1971; Barlow et al., 1987).

298

H. Laburn et al.

Fig. 17.2. Fetal lamb (F), and pregnant ewe (M) body temperatures as measured by radiotelemetry in a group of seven animals over the last 35 days of pregnancy (lambing at day 0). Lowest panel shows the feto-maternal (F–M) body temperature gradient. All points are means ± standard error of a mean. Reprinted with permission from the American Physiological Society, from Laburn et al. (1992).

The Thermal Physiology of the Ruminant Fetus

299

Fig. 17.3. Top panel shows circadian variations in fetal lamb and pregnant ewe body temperatures over a 24-h period, and lower panels show changes in the feto-maternal (F-M) temperature gradient, and feto-amniotic (F-A) temperature gradient. The F-A gradient is approximately midway between the F-M gradient, confirming the amnion as a conduit of heat between fetus and maternal tissues. Other details as in Fig. 17.1. Reprinted with permission from the American Physiological Society, from Laburn et al. (1992).

Fetal hyperthermia A potentially serious consequence of the fetus having a higher body temperature than that of its mother, and of being thermally clamped to its mother, is that the fetus experiences greater hyperthermia than the mother does, when maternal body temperature rises. Heat is damaging, especially to developing tissues, and the developing central nervous system is particularly susceptible. A rise in temperature of 1.5°C arrests fetal brain cell division, and a 3°C rise kills dividing cells (Gericke et al., 1989). As a result, maternal hyperthermia in early pregnancy leads to congenital abnormalities (Edwards et al., 1995; Chambers et al., 1998). In later pregnancy, raised intrauterine temperature is

300

H. Laburn et al.

Fig. 17.4. Body temperatures of a pregnant ewe (M) and her fetal lamb (F). Measurements were made while the ewe was housed in laboratory conditions (a), and in field conditions (see text, b). Data for a 3-day (a) or a 5-day period (b) are shown.

associated with fetal growth retardation, such that the lambs of animals kept at abnormally high ambient temperatures during their pregnancy have significantly lower body mass than their counterparts whose mothers were kept in more moderate conditions (Alexander and Williams, 1971; Dreiling et al., 1991).

The Thermal Physiology of the Ruminant Fetus

301

Fetal body temperature during maternal hyperthermia and hypothermia The fetus apparently has no way of controlling its body temperature independently of that of its mother, nor can the fetus escape the thermal influence of its immediate environment, that is, maternal tissue. Thus, irrespective of the prevailing levels of fetal heat production and heat loss, fetal body temperature rises whenever maternal body temperature does. Rises in maternal body temperature occur in ruminants, for example, during exposure to hot environments and when animals exercise. We have raised pregnant ewe body temperature in three different experimental protocols, and have examined fetal lamb body temperature in those circumstances (Laburn et al., 1992). Figure 17.5 summarizes our results. Exposure of the ewes to a hot environment for 2 h (Fig. 17.5a) or to 30 min of treadmill exercise (Fig. 17.5b) caused their body temperature to rise, and fetal lamb temperature rose too. However, we observed that the fetal lamb temperature rise did not exactly parallel the ewe’s temperature. Instead, the difference between fetal and maternal temperature fell significantly. The result was that, in these situations of raised ewe body temperature, fetal body temperature was prevented from rising to the same extent by a reduction in the F-M gradient, and a consequent reduction in the risk of hyperthermia for the fetus. In a single experiment on a pregnant baboon, external heat exposure reduced the F-M gradient in that species too (Morishima et al., 1975), and we recently have found a similar response in goats (unpublished oberservations). Given the apparent ubiquity of a reduction in the F-M gradient during maternal ambient heat stress and during exercise in sheep (we also have observed the phenomenon during labour, see below), we have asked ourselves whether there could be a common mechanism for reducing the risk of fetal hyperthermia. Increased blood flow in umbilical or uterine blood vessels could account for the significant fall in the F-M gradient. Intuitively, one might expect a decrease in perfusion of visceral vascular beds in heat-stressed animals, to divert blood flow to the periphery. Indeed there are several reports claiming that uterine blood flow does decrease in pregnant ewes exposed to heat (Alexander et al., 1987) and exercise (Bell et al., 1984), although the effects do not occur in pregnant women (Vähä-Eskeli et al., 1991). If uterine blood flow fell, then the F-M thermal gradient should have increased, but the researchers who measured uterine flow did not measure fetal temperature. Evidence in support of our hypothesis is that from Cefalo and Hellegers (1978), who showed that uterine blood flow rose during ewe hyperthermia, diminishing only after maternal body temperature had risen by more than 2.0°C, and that umbilical blood flow also increased as ewe body temperature rose, falling only after ewe temperature had risen by 2.5°C. Walker et al. (1995) demonstrated a rise in blood flow to many tissues in fetal lambs during maternal heat exposure, and a consequent rise in vascular conductance, which would increase heat loss from the fetus via placental and non-placental routes. Thus, at moderate levels of maternal hyperthermia, both mother and fetal animal may activate mechanisms to reduce the risk of fetal hyperthermia. Under more severe heat stress, the mother may attend to her own thermoregulatory needs, even though that means compromising her fetus. Fever is the specific form of hyperthermia which accompanies inflammation and infection, and in our third hyperthermic protocol, we examined the consequences for fetal lamb body temperature when the pregnant ewe is febrile (Laburn et al., 1992). Fig. 17.5c shows the typical fever response when ewes were injected with the purified

302

H. Laburn et al.

Fig. 17.5. Summary of changes in fetal lamb (F) and pregnant ewe (M) body temperatures and effects on the feto-maternal (F-M) gradient of exposure of the ewes to various conditions of thermal stress. (a) exposure at rest to 40°C and 60% relative humidity. (b) Treadmill exercise (2 km h⫺1, 5° gradient) for 30 min at room temperature. (c) Experimental fever following intravenous injection of bacterial lipopolysaccharide into the ewe at time zero. (d) Exposure to 4°C. Reprinted with permission from the American Physiological Society, from Laburn (1996).

The Thermal Physiology of the Ruminant Fetus

303

lipopolysaccharide of a Gram-negative bacterium. Ewe body core temperature rose after the pyrogen injection and, as with the other forms of hyperthermia, body temperature of the fetus rose also, but during maternal fever, contrary to what occurred in the other hyperthermias, the F-M gradient increased significantly. The rise in temperature therefore was significantly greater for the fetus than it was for the mother animal, during maternal fever. For hyperthermias associated with increased thermal loads, of external (hot environments) or internal (exercise) origin, the mother’s physiological response was to increase her heat loss, using peripheral vasodilation and/or evaporation to dissipate heat. During fever, the body reacts as if it were exposed to cold to achieve the elevated temperature. Typically, animals vasoconstrict and shiver. We have hypothesized that during the febrile episode, vasoconstriction in the uterine circulation, as well as in other peripheral vascular beds, compromised fetal heat loss. The thermal consequences of fever in pregnant ruminants, of whatever cause, therefore may be a significant risk to fetal well-being. Peripheral vasoconstriction also occurs in response to a lowering of body temperature, and we suspect that, in a way similar to that which we postulate occurred during the febrile episode, pregnant ewes exposed to cold ambient conditions reduce blood flow in the uterine circulation, and fetal heat loss is decreased. In the case of maternal hypothermia, however, the resulting rise in F-M gradient, attenuates the fall in fetal body temperature, a response which benefits the fetus. The effect is shown in Fig. 17.5d, and also has been observed by Morishima et al. (1975) in a baboon. Thus, apart from the specific case of maternal fever, maternal or fetal mechanisms involving changes in blood flow conspire to reduce changes in fetal body temperature when the mother is subjected to thermal insult.

Birth-related changes in body temperatures Changes in body temperature of the ewe and fetal lamb There are three phases of body temperature change accompanying parturition. The first is that which occurs prior to lambing, in which we have observed that ewe body temperature falls by a few tenths of a degree over the few days leading up to labour (Fig. 17.2). This phenomenon has been observed in horses (Shaw et al., 1988; Haluska and Wilkins, 1989), cattle (Ewbank, 1963), dogs (Concannon et al., 1977) and sheep (Ewbank, 1969; Laburn et al., 1992), and would be a useful predictor of parturition if daily body temperature measurements were made routinely in ruminants. It may have to do with the fall in progesterone plasma concentration as labour becomes imminent (Janowski et al., 1995). The second phase is that of the exertion of labour, and myometrial contractions, which produce sufficient heat in the active tissues (Gemzell et al., 1957; Marx and Loew, 1975) to cause body temperature to rise. Fetal temperature inevitably rises too (Fig. 17.6), but is accompanied by a reduced F-M thermal gradient (Laburn et al., 1994), as it is in maternal locomotor exercise. The effect of attenuating the F-M gradient during labour may have to do with reducing any increase in body temperature in the fetal/neonatal animal. The lower the body temperature during the process of birth,

304

H. Laburn et al.

Fig. 17.6. Changes in body temperatures of fetus or lamb, and mother animal, and the feto-maternal (F-M) temperature gradient, measured for 2 h before and 4 h after lambing (at time zero). Each point is the mean ± standard error of a mean (SEM) of seven ewes and eight fetuses/lambs. Reprinted with permission from Birkhäuser Verlag AG, from Laburn et al. (1994).

the lower the risk of ischaemic brain damage, should neonatal circulation or respiration be compromised in the birth process (Kuroiwa et al., 1990). The third phase of the thermal response to parturition is that more dramatic

The Thermal Physiology of the Ruminant Fetus

305

phase, for the newborn ruminant animal at least, immediately after birth. The measurements shown in Fig. 17.6 were taken in a group of ewes giving birth in indoor conditions. The neonatal lamb, delivered soaked in amniotic fluid, is confronted by an environment which, compared with the intrauterine environment, is cold and dry. Evaporation of fluid from the lamb’s skin, and its predisposition to heat loss by virtue of its high surface area to mass ratio, results in a precipitous fall in body temperature, by about 1.5°C within minutes of birth. Lambs in the field are delivered into much colder conditions (Barlow et al., 1987) so one should expect that the fall in neonatal body temperature would be more precipitous than it was in our experiments. Nevertheless, within an hour, in the indoor environment at least, thermoregulatory effector mechanisms come into play, not only to halt the plummeting body temperature, but to reverse it. By 3 h post-partum, lamb and ewe body temperatures are not significantly different (Fig. 17.6).

Development of thermoregulatory mechanisms The remarkable ability of neonatal animals to cope with the environment into which they are born is the result, in large part, of their ability, very soon after birth in ruminant neonates, to activate non-shivering thermogenesis (NST) in the brown adipose tissue with which they are well endowed at birth (Nedergaard and Cannon, 1992). NST appears when an uncoupling protein, in brown adipose cells, causes short-circuiting of protons in the process of oxidative phosphorylation (Cannon and Nedergaard, 1985) such that extra heat, rather than adenosine triphosphate, is produced. In sheep, the concentration of mRNA for uncoupling protein peaks just before birth (Casteilla et al., 1989). Cold exposure stimulates beta-adrenergic activity and in turn receptors on the brown adipose cells, and synthesis of the uncoupling protein. Because of the large blood supply to brown adipose tissue, heat produced is transported from its source to other body sites, and body temperature of the neonate rises, unless the ambient thermal conditions and heat loss from the lamb overcome the capacity for NST (Barlow et al., 1987). The ability to activate NST within minutes of birth is crucial to the survival of the neonatal lamb. How does the fetal lamb acquire the mechanisms for activation of NST, when the necessary stimuli for inducing the development of the mechanisms are absent in utero? The late Tania Gunn and her colleagues (Gunn and Gluckman, 1983; Gunn et al., 1991) carried out elegant experiments in which they were able to simulate the exposure to cold, and other features that the neonate encounters at birth, but in fetal lambs, in utero. They measured plasma glycerol concentration and the temperature in brown adipose tissue as indicators of the activation of NST (Fig. 17.7). Neither oxygenation nor sympathetic stimulation alone, nor jointly, could activate NST. What was necessary was stimulation of cold receptors in the fetal lamb, and a simulation of cutting the umbilical cord. Thus, NST is possible in the fetal lamb, but the intrauterine thermal environment, and an inhibitor of the response emanating from the placenta, possibly prostaglandins (Andrianakis et al., 1989; Gunn et al., 1993), appear to suppress NST in utero. Shivering too can be induced experimentally in certain circumstances, in fetal lambs (Dawes, 1968) and so can vasoconstriction in response to a cold stimulus (Gunn

306

H. Laburn et al.

Fig. 17.7. Data from experiments performed by Gunn et al. (1991) showing the activation of non-shivering thermogenesis (NST) in a group of fetal lambs during 5 h of simulated post-birth conditions. Top curve, changes in plasma glycerol concentrations. Bottom curve, increase in brown adipose tissue (BAT) temperature above that of the lamb’s core body temperature. Lambs were ventilated, and primed with a ␤-stimulant. NST was stimulated by cooling the fetus in utero, but only after the umbilical cord was snared. Reprinted with permission from the American Physiological Society, from Laburn (1996).

and Gluckman, 1983). Normally, however, fetal lambs probably are peripherally vasodilated (Schröder et al., 1988) and presumably shivering, like NST, is suppressed in utero. Other thermoregulatory mechanisms, including the ability to pant in response

The Thermal Physiology of the Ruminant Fetus

307

to high ambient conditions, also probably are mature in the late-gestation fetal lamb, but presumably are suppressed.

Post-partum body temperatures The body temperatures of both neonatal lamb and its mother remain unexpectedly high in the days and weeks following parturition. Figure 17.8 shows that at constant ambient temperature, neonatal body temperaure peaks at about 7 days post-partum, and then declines towards adult levels by a month of age. In the post-partum ewe, lactation appears to promote a sustained elevation of body temperature. A comparison between Figs 17.6 and 17.8 shows that the body temperature of the lactating ewe is

Fig. 17.8. Changes in fetal and lamb (upper panel) and ewe (bottom panel) body temperatures measured for 10 days prior to lambing and for 7 weeks thereafter. Each point is the mean ± standard error of a mean for seven ewes and eight fetuses/lambs. Lambing is indicated by a vertical dashed line.

308

H. Laburn et al.

higher even than that towards the end of pregnancy, at least for 3 weeks. What the mechanism is of the sustained rise in body temperature is unknown, but it may be a defended elevation in temperature; lactating ewes have attenuated fever responses in response to administration of a Gram-negative pyrogen (S. Glassom et al., South Africa, unpublished observations).

Preterm labour and abortion There is anecdotal evidence that acute ambient heat or cold stress may precipitate preterm labour in ruminants. Infection, too, can cause premature delivery, and in this instance the cause appears to be fetal death. We simulated maternal, fetal and intrauterine infection by injecting Gram-positive and Gram-negative pyrogens directly into the pregnant ewe, its fetus and amniotic cavity. Even though the dose of pyrogen administered to adult or fetal animal was not sufficient to lead to severe fetal hyperthermia, the pyrogens had potentially serious effects on fetal well-being. Table 17.1 shows the percentage of abortions we observed after administering Gram-positive or Gram-negative pyrogens to ewe, fetus, or into the amniotic fluid. The presence of bacterial material (even killed organisms or purified extracts) induced abortion in a proportion of sheep, with a greater risk of abortion when the pyrogen was present in the fetus than in the mother animal, and the highest risk being associated with the presence of pyrogenic material in the amniotic cavity. It is likely that cytokines, released from fetal and/or placental tissue (Mitchell et al., 1991; Gibbs et al., 1992) as a result of the pyrogen presence, induced inflammatory changes in the placenta, which led to severely compromised fetal function, and ultimately fetal death.

Acknowledgements The work reported in this article was supported by the Council of the University of the Witwatersrand, Johannesburg, and the South African Medical Research Council.

Table 17.1. Abortion following pyrogen injection. Percentage of abortions, and time to abortion (in days, mean ± standard error of a mean (SEM), after either Gram-positive or Gram-negative pyrogens were injected directly into the circulation of pregnant ewes (n = 12), or fetal animals (n = 14), or into the amniotic fluid (n = 4).

Pyrogen to ewe Pyrogen to fetus Pyrogen to amnion

Abortion (%)

Days to abortion ± SEM

41 64 75

10.5 ± 2.6 4.6 ± 0.9 1.6 ± 0.7

The Thermal Physiology of the Ruminant Fetus

309

References Abrams, R., Caton, D., Clapp, J. and Barron, D.H. (1970) Thermal and metabolic features of life in utero. Clinical Obstetrics and Gynecology 13, 549–564. Adamsons, K. and Towell, M.E. (1965) Thermal homeostasis in the fetus and newborn. Anesthesiology 26, 531–548. Alexander, G. and Williams, D. (1971) Heat stress and development of the conceptus in domestic sheep. Journal of Agricultural Science, Cambridge 76, 53–72. Alexander, G., Hales, J.R.S., Stevens, D. and Donnelly, J.B. (1987) Effects of acute and prolonged exposure to heat on regional blood flows in pregnant sheep. Journal of Developmental Physiology 9, 1–15. Andrianakis, P., Walker, D.W., Ralph, M.M. and Thorburn, G.D. (1989) Effect of inhibiting prostaglandin synthesis in pregnant sheep with 4-aminoantipyrine under normothermic and hyperthermic conditions. American Journal of Obstetrics and Gynecology 161, 241–247. Assali, N.S. and Westin, B. (1962) Effects of hypothermia on uterine circulation and on the fetus. Proceedings of the Society for Experimental Biology and Medicine 109, 485–488. Barlow, R.M., Gardiner, A.C., Angus, K.W., Gilmour, J.S., Mellor, D.J., Cuthbertson, J.K., Newlands, G. and Thompson, R. (1987) Clinical, biochemical and pathological study of perinatal lambs in a commercial flock. Veterinary Record 120, 357–362. Bell, A.W., Hales, J.R.S., Fawcett, A.A. and King, R.B. (1984) Contribution of the uterus to circulatory requirements of exercise and heat stress in pregnant sheep. In: Hales, J.R.S. (ed.) Thermal Physiology. Raven Press, New York, pp. 377–380. Bell, A.W., Battaglia, F.C. and Meschia, G. (1987) Relation between metabolic rate and body size in the ovine fetus. Journal of Nutrition 117, 1181–1186. Cannon, B. and Nedergaard, J. (1985) The biochemistry of an inefficient tissue: Brown adipose tissue. Essays in Biochemistry 20, 110–164. Casteilla, L., Champigny, O., Bouilland, F., Robelin, J. and Ricquier, D. (1989) Sequential changes in the expression of mitochondrial protein mRNA during the development of brown adipose tissue in bovine and ovine species. Biochemical Journal 257, 665–671. Cefalo, R.C. and Hellegers, A.E. (1978) The effects of maternal hyperthermia on maternal and fetal cardiovascular and respiratory function. American Journal of Obstetrics and Gynecology 131, 687–694. Chambers, C.D., Johnson, K.A., Dick, L.M., Felix, R.J. and Jones, K.L. (1998) Maternal fever and birth outcome: a prospective study. Teratology 58, 251–257. Concannon, P.W., Powers, M.E., Holder, W. and Hansel, W. (1977) Pregnancy and parturition in the bitch. Biology of Reproduction 16, 517–526. Dawes, G.S. (1968) Fetal and Neonatal Physiology. Year Book Medical Publishers Inc., Chicago, pp. 191–209. Dreiling, C.E., Carman, F.S. and Brown, D.E. (1991) Maternal endocrine and fetal metabolic responses to heat stress. Journal of Dairy Science 74, 312–327. Edwards, M.J., Shiota, K., Smith, M.S.R. and Walsh, D.A. (1995) Hyperthermia and birth defects. Reproductive Toxicology 9, 411–425. Ewbank, R. (1963) Predicting the time of parturition in the normal cow. The Veterinary Record 75, 367–371. Ewbank, R. (1969) The fall in rectal temperature seen before parturition in sheep. Journal of Reproduction and Fertility 19, 569–571. Gemzell, C.A., Robbe, H., Stern, B. and Ström, G. (1957) Observations on circulatory changes and muscular work in normal labour. Acta Obstetrica et Gynecologica Scandinavica 36, 75–92 Gericke, G.S., Hofmeyr, G.J., Laburn, H.P. and Isaacs, H. (1989) Does heat damage fetuses? Medical Hypotheses 29, 275–278.

310

H. Laburn et al. Gibbs, R.S., Romera, R., Hillier, S.L., Eschenbach, D.A. and Sweet, R.L. (1992) A review of premature birth and subclinical infection. American Journal of Obstetrics and Gynecology 166, 1515–1528. Gilbert, R.D., Schröder, H., Kawamura, T., Dale, P.S. and Power, G.G. (1985) Heat transfer pathways between fetal lamb and ewe. Journal of Applied Physiology 59(2), 634–638. Gunn, T.R. and Gluckman, P.D. (1983) Development of temperature regulation in the fetal sheep. Journal of Developmental Physiology 5, 167–179. Gunn, T.R., Ball, K.T., Power, G.G. and Gluckman, P.D. (1991) Factors influencing the initiation of nonshivering thermogenesis. American Journal of Obstetrics and Gynecology 164, 210–217. Gunn, T.R., Ball, K.T. and Gluckman, P.D. (1993) Withdrawal of placental prostaglandins permits thermogenic responses in fetal sheep brown adipose tissue. Journal of Applied Physiology 74(3), 998–1004. Haluska, G.J. and Wilkins, K. (1989) Predictive utility of pre-partum temperature changes in the mare. Equine Veterinary Journal 21, 116–118. Hart, F.M. and Faber, J.J. (1965) Fetal and maternal temperatures in rabbits. Journal of Applied Physiology 20, 737–741. Janowski, T., Zdunczyk, S., Podhalicz-Dziegielewska, M., Ras, A. and Chmielewski, A. (1995) Course of pregnancy, steroid hormone levels and maturation process of placentomes after oestrone infusion in cows near term. Zentralbl Veterinarmed A 42(5), 345–350. Kuroiwa, T., Bonnekoh, P. and Hossmann, K.-A. (1990) Prevention of postischemic hyperthermia prevents ischemic injury of CA1 neurons in gerbils. Journal of Cerebral Blood Flow and Metabolism 10, 550–556. Laburn, H.P. (1996) How does the fetus cope with thermal challenges? News in Physiological Sciences 11, 96–100. Laburn, H.P., Mitchell, D. and Goelst, K. (1992) Fetal and maternal body temperatures measured by radiotelemetry in near-term sheep during thermal stress. Journal of Applied Physiology 72(3), 894–900. Laburn, H.P., Goelst, K. and Mitchell, D. (1994) Body temperatures of lambs and their mothers measured by radio-telemetry during parturition. Experientia 50, 708–711. Marx, G.F. and Loew, D.A.Y. (1975) Tympanic temperature during labour and parturition. British Journal of Anaesthesia 47, 600–602. Mitchell, M.D., Branch, D.W., Lundin-Schiller, S., Romero, R.J., Daynes, R.A. and Dudley, D.J. (1991) Immunologic aspects of preterm labor. Seminars in Perinatology 15, 210–224. Morishima, H.O., Glaser, B., Niemann, W.H. and James, L.S. (1975) Increased uterine activity and fetal deterioration during maternal hyperthermia. American Journal of Obstetrics and Gynecology 121, 531–538. Nedergaard, J. and Cannon, B. (1992) Brown adipose tissue: development and function. In: Polin, R.A. and Fox, W.W. (eds) Fetal and Neonatal Physiology. Saunders, Philadelphia. Schröder, H., Gilbert, R.D. and Power, G.G. (1988) Computer model of fetal-maternal heat exchange in sheep. Journal of Applied Physiology 65(1), 460–468. Shaw, E.B., Houpt, K.A. and Holmes, D.F. (1988) Body temperature and behaviour of mares during the last two weeks of pregnancy. Equine Veterinary Journal 20, 199–202. Vähä-Eskeli, K., Pirhonen, J., Seppänen, A. and Erkkola, R. (1991) Doppler flow measurement of uterine and umbilical arteries in heat stress during late pregnancy. American Journal of Perinatology 8, 385–389. Walker, D.W., Hales, J.R.S., Fawcett, A.A. and Pratt, N.M. (1995) Cardiovascular response to heat stress in late gestation fetal sheep. Experimental Physiology 80, 755–766. Wood, C. and Beard, R.W. (1964) Temperature of the human foetus. Journal of Obstetrics and Gynaecology of the British Commonwealth 71, 768–769.

18

Regulation of Nutrient Partitioning During Lactation: Homeostasis and Homeorhesis Revisited

D.E. BAUMAN Department of Animal Science, Cornell University, Ithaca, New York, USA

Introduction Nature has accorded a high priority to lactation, and the ability of mammals to synthesize milk is essential for survival of the newborn. The mammary glands have a high metabolic rate during lactation, yet this organ is unique because its biosynthetic products represent no direct benefit to the mother. Rather, the extensive rate of nutrient use to make milk imposes a substantial demand on the mother and mandates that the metabolism of the mammary glands and other maternal tissues be coordinated. Thus, during lactation many physiological changes occur and there are profound alterations in the metabolism of many tissues. The extensive physiological adaptations which occur during lactation have the overall effect of providing the proper quantity and pattern of nutrients for milk synthesis. Thus, lactation provides an excellent opportunity to elucidate the broad concepts of metabolic regulation and identify specific mechanisms involved in the partitioning of nutrients. The regulation of nutrient use is not only important for a successful lactation, but it also represents the physiological basis for differences in productive efficiency, and this is of special importance in agricultural species. Furthermore, if the coordination of nutrient use is inadequate then animal well-being is compromised which may result in stress, subclinical conditions and metabolic disorders. In 1980, Currie and I reviewed the regulation of nutrient partitioning during pregnancy and lactation. In discussing the concepts of regulation, we crystallized the concept of homeorhesis, proposed mechanisms, and addressed the interrelations between homeostasis and homeorhesis (Bauman and Currie, 1980). This review will elaborate and trace the development of the concepts of homeostasis and homeorhesis, focus on the regulation of nutrient use during lactation, and review current understanding of the mechanisms and applications of these concepts. Because of its breadth, reviews will be cited frequently for various aspects of the biology.

© CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

311

312

D.E. Bauman

Concepts of regulation At 12 minutes past eight on 3 February 1783, Lavoisier initiated experiments with a guinea pig in what is now recognized as the first animal calorimeter (Blaxter, 1989; Welch, 1991). In collaboration with Laplace, he carried out two experiments in a 24 h period, which for the first time firmly linked the evolution of heat by animals with the consumption of oxygen and the formation of carbon dioxide. These simple, but elegant, experiments established that life is a chemical process and represented an initial important step in the search to understand bioenergetics and the regulation of metabolism. It was almost 100 years later that another milestone occurred when Bernard, a physiologist, recognized the ability of living organisms to maintain their own constancy. He cogently observed that ‘all the vital mechanisms, however varied they may be, have only one object, that of preserving constant the conditions of life in the internal environment’ (Bernard, 1878). Using Bernard’s concept of a stable ‘milieu interieur’ as the cornerstone, Cannon concluded that the coordinated physiological reactions which maintain most of the steady states in a living organism were complex, involving many different tissues and organs ‘all working cooperatively’ (Cannon, 1932). He recognized that higher organisms had more elaborate and effective systems to maintain steady-state conditions, and suggested a special designation for these states – homeostasis. Cannon (1929) chose the term homeostasis with a great deal of thought and foresight. He concluded that the prefix ‘homeo’ was preferred because it meant ‘like’ or ‘similar’, thereby indicating some variation. He rejected the prefix ‘homo’ because it meant ‘same’ and implied a rigid constancy. Likewise Cannon (1929) chose ‘stasis’ because it represented a ‘condition’. By choice of terminology and examples to elaborate the concept, Cannon (1929, 1932) emphasized that dynamic regulation and coordination were key features of homeostasis. Today the concept of homeostasis is well known to biologists, and there are many systems where the positive and negative feedback controls to preserve steady state are well established. Glucose was an example used by Bernard (1878) in developing the concept of milieu interieur and by Cannon (1932) in crystallizing the concept of homeostasis. The homeostatic controls to maintain steady-state conditions for glucose are also of special significance during lactation. Glucose is critical during lactation because its uptake by the mammary gland is essential for the synthesis of milk lactose, the major osmotic regulator of milk volume. The pancreatic hormones, insulin and glucagon, are key controls of glucose homeostasis. Thus, acute regulation of plasma glucose concentration by the reciprocal actions of insulin and glucagon ensure the proper balance in glucose supply and utilization by body tissues and organs during lactation. While the concept of homeostasis is universally accepted today, it resulted in substantial debate over several decades following its introduction (Waddington, 1942, 1953; Lerner, 1954; Lewontin, 1956). In part, this debate focused on the fact that the concept was not adequate to address regulation over a wide range of physiological and developmental situations. Waddington (1957) reasoned that there are at least three types of temporal change in biological adaptations, all occurring simultaneously. The first, on the longest time scale, is evolution. The second, of medium time scale relates to development throughout the life cycle and the third, on the shortest time scale,

Regulation of Nutrient Partitioning During Lactation

313

applies to the day-to-day events of physiology. Thus, physiological changes do more than preserve a day-to-day constant state of animal well-being, they are also altered as time passes which produces the longer-term changes associated with ontogeny. To illustrate this, Waddington (1957) compared biological processes in an animal with a musical composition. The short-scale physiology of the animal would be similar to the vibrations of individual notes. The medium-scale life-history would be analogous to the melodic phrases into which the notes build themselves; and the animal’s long-scale evolution would represent the structure of the whole musical composition, in which the melodies are repeated and varied. In the three decades following Cannon’s crystallization of homeostasis, several terms were introduced to accommodate the developmental aspect of regulation. Waddington played a central role in this discussion, and introduced the terms ‘developmental homeostasis’ and ‘canalization’ (Waddington, 1942, 1953). Later, he suggested the term ‘homeorhesis’ by analogy to homeostasis, and related it to describe ‘flow’ to a new equilibrium or pathway of developmental change over the life cycle (Waddington, 1957). In reviewing this early debate, one cannot help but marvel at the creativeness and originality of Waddington and others, bearing in mind that this was a period prior to discovery of DNA and the genetic code, prior to the identification of second message cascades and signal transduction systems, and prior to the development of analytical techniques and molecular probes for measuring proteins and gene transcripts. Obviously, some aspects of this early discussion are incorrect as shown by subsequent discoveries, and it is often difficult to relate some of the discussion to current knowledge of biology. Nevertheless, the reasoning and insight of these early pioneers in biological regulation are impressive. In 1980, Currie and I crystallized the concept of homeorhesis in the regulation of nutrient partitioning and physiological processes. We defined homeorhesis as ‘the orchestrated or coordinated changes in metabolism of body tissues necessary to support a physiological state’ (Bauman and Currie, 1980). The Greek derivation of homeorhesis is ‘uniform-flow’. Thus, homeorhetic regulation involves a coordination of metabolism to ensure a uniform flow of nutrients in support of a physiological state. Of particular importance, we characterized key features of homeorhetic control: its chronic nature, i.e. hours or days versus the seconds or minutes required for most examples of homeostatic regulation, its simultaneous influence on multiple tissues and systems with apparently unrelated functions, and its mediation through altered responses to homeostatic signals (Bauman and Currie, 1980; Bauman, 1984). The physiological states of lactation and pregnancy were used as the initial examples to illustrate the concept of homeorhesis, and we later extended it to regulation during growth (Bauman et al., 1982). Kennedy (1967) also applied the concept of homeorhesis to growth in an earlier review which detailed the changes in the homeostatic regulation of food intake that occur at different stages of development. Others have extended this concept to a broad array of physiological processes, and in some cases other terms have been introduced which, for all practical purposes, are synonymous and/or consistent with the concept of homeorhesis. These have included homeorheusis (Nicoläidis, 1983), teleophoresis (Chilliard, 1986), rheostasis (Mrosovsky, 1990) and poikilostasis (Kuenzel et al., 1999). For example, Mrosovsky (1990) introduced the term rheostasis (rheo = change) and defined it as ‘a condition in which, at any one instant, homeostatic defenses are present but over a span of time, there is a change in

314

D.E. Bauman

the level that is defended’. Likewise, Kuenzel et al. (1999) introduced the term poikilostasis (poikilo = various) to describe the dynamic shifts in homeostatic regulation of metabolism and food intake that occur in birds over different physiological states such as growth, migration and reproduction. In recognition of the original work of Waddington (1957) and the earlier use by Kennedy (1967), we chose to use the term homeorhesis (Bauman and Currie, 1980). Nevertheless, in all cases the concept relates to the ability of the animal to adjust biological processes in a manner to support a dominant physiological state for animal wellbeing and survival of the species. As illustrated in Table 18.1, the general concept represented by homeorhesis has been extended to an impressive range of biological situations encompassing many different physiological, nutritional and even pathological states. In particular, Mrosovsky (1990) provided an extensive list of examples of this regulation in terms of both physiological states and biological processes. A key physiological state in the survival of mammals is lactation, and the remainder of this review will relate the concepts of regulation to lactation.

Adaptations to lactation Lactation represents an impressive example of homeorhesis, and at peak lactation the proportion of nutrients used by the mammary gland can be extraordinary. For example, in high-producing dairy cows nutrient utilization by the mammary glands exceeds that of the rest of the body causing Brown (1969) to suggest that the cow should be envisioned as an appendage to the udder rather than vice versa. Nutrient supply and use during lactation are similar for all species, although there are qualitative and quantitative differences (Linzell, 1967; Williamson et al., 1995). In all species, the extent of nutrient use for milk synthesis requires integrated regulation of the metabolism of the mammary glands and other body tissues. Using the dairy cow as an example, a partial list of the physiological adaptations which occur during lactation is presented in Table 18.2. The importance accorded to lactation in mammals is demonstrated by the fact that the physiological adaptations involve many, perhaps most, of the body tissues and relate to the metabolism of all nutrient classes. The net effect is that the increase in mammary gland metabolic rate and nutrient use which occurs during lactation coincides with alterations in the metab-

Table 18.1. Partial list of physiological situations where the general concept representing homeorhetic regulation has been applied.a Lactation Pregnancy Growth Puberty Ageing Chronic undernutrition Chronic illness a References

Hibernation Premigration/migration Egg laying Incubation anorexia Seasonal cycles Exercise

include Kennedy (1967), Bauman and Currie (1980), Dilman (1982), Mrosovsky (1990), Wade and Schneider (1992), Vernon (1998), and Kuenzel et al. (1999).

Regulation of Nutrient Partitioning During Lactation

315

Table 18.2. Partial list of physiological adaptations which occur in lactating dairy cows. Process or tissue

Response

Mammary tissue

Increased number of secretory cells Increased nutrient use Increased supply of blood

Food intake

Increased quantity

Digestive tract

Increased size Increased absorptive capacity Increased rates of nutrient absorption

Liver

Increased size Increased rates of gluconeogenesis Increased glycogen mobilization Increased protein synthesis

Adipose tissue

Decreased de novo fat synthesis Decreased uptake of preformed fatty acids Decreased re-esterification of fatty acids Increased lipolysis

Skeletal muscle

Decreased glucose utilization Decreased protein synthesis Increased protein degradation

Bone

Increased mobilization of Ca and P

Heart

Increased cardiac output with a larger percentage going to the mammary glands

Plasma hormones

Decreased insulin Increased somatotropin Increased prolactin Increased glucocorticoids Decreased thyroid hormones Decreased IGF-I

olism of other body tissues so that an adequate quantity and pattern of nutrients to support milk synthesis is ensured. Several reviews have detailed the physiological adaptations which occur during lactation and readers are referred to these for quantitative information (Williamson, 1980; Bauman and Elliot, 1983; Chilliard, 1986, 1987; Vernon, 1989; Williamson and Lund, 1994; McNamara, 1995). Nevertheless, some of these adaptations merit mention as examples which illustrate the concepts of homeostasis and homeorhesis. During lactation, food intake is increased in many species. Corresponding adaptations also occur in the size and absorptive capacity of the gastrointestinal tract, thereby allowing for an increased absorption of nutrients (Bauman and Elliot, 1983; Vernon, 1989). In rodents where the milk demand of the nursing pups gradually increases, the increase in food intake is also gradual and can amount to a 300–400% increase

316

D.E. Bauman

compared to the non-lactating state (Cripps and Williams, 1975). Furthermore, the magnitude of the increase in intake is related to the number of nursing pups (Millar, 1979), providing convincing evidence of the coordinated nature of these changes. Increases in feed intake and digestive tract size also occur in dairy cows, with the magnitude of the increase in voluntary intake being related to milk yield (Bauman and Elliot, 1983). However, for this species the increase in feed intake occurs over a longer interval so that high-producing dairy cows may not achieve a positive energy balance until 8–12 weeks post-partum (Bauman and Currie, 1980). Despite adjustments in food intake, many species rely on body reserves during early lactation. Vernon (1998) pointed out that this results in many body tissues having metabolic characteristics that are typical of the adaptations which occur with chronic undernutrition. However, the extent to which reserves are needed in early lactation to meet nutrient requirements, especially energy requirements, varies among species. In some species, the use of energy reserves is minimal to modest because either the amount of milk required by the nursing young is limited (e.g. human and guinea pig), or the increase in voluntary intake is nearly adequate to meet requirements (e.g. rat and mouse). In other species such as the cow, goat and pig, the use of body reserves is more extensive during the lactation cycle. For example, lactational yields in dairy cows are related to the magnitude of body reserve utilization (Bauman et al., 1985), and in highproducing cows the mobilization of body fat during the first month of lactation can be energetically equivalent to over one-third of the milk produced (Bauman and Currie, 1980). The use of body reserves is extraordinary in several species and some examples of these merit special mention because they dramatically illustrate the concepts of metabolic regulation. Seals make spectacular use of body reserves to meet their nutrient requirements during lactation (Riedman, 1990; Oftedal, 1993). This is illustrated by the pregnant elephant seal which gives birth to a single pup and sustains a 28 day lactation. The nursing pup averages a rate of body weight gain of approximately 10% day⫺1. In contrast, the mother loses weight, as she neither eats nor drinks throughout the 4 week lactation. Maternal use of body reserves over the lactation interval results in a 42% loss of body weight, and this represents a 58% reduction in body fat content and a 14% reduction in body lean weight (Costa et al., 1986). Whales, especially baleen whales, also make extraordinary use of body reserves during lactation and the blue whale provides an example (Lockyer, 1981; Oftedal, 1993). Female blue whales have a body weight of about 80,000 kg and during pregnancy their 40,000–55,000 kg gain in body weight primarily represents the accretion of body reserves. The blue whale calf weighs about 2500 kg at birth and gains at the rate of 80–100 kg day⫺1 during the 7 month lactation. During this interval the blue whale mother produces about 90 kg of milk day⫺1. Even more impressive, she relies almost exclusively on body reserves to support the nutrient needs for milk synthesis and her own sustenance, as she eats little, if at all, throughout the 7 month lactation. In seals and whales, as well as other lactating mammals, the physiological adaptations to support lactation are extensive. Based on work with laboratory and farm animals, we can envision how body fat reserves may serve to meet the energy requirements of the lactating elephant seal and blue whale. However, body reserves in these species must also account for the protein, carbohydrate, mineral and vitamin components of milk. To a large extent the qualitative and quantitative regulation of

Regulation of Nutrient Partitioning During Lactation

317

maternal metabolism of these reserves is unknown. Nevertheless, the regulation of nutrient utilization involves homeostatic controls operating on a minute-by-minute basis to maintain constant conditions every day of the mother’s life, and at the same time homeorhetic controls must function during lactation to orchestrate body processes in such a manner to partition the appropriate pattern and quantity of nutrients to support milk synthesis. Furthermore, the success of these regulatory processes is essential to ensure well-being of the lactating mother and survival of the nursing young.

Mechanisms The metabolic adaptations occurring with the onset of lactation are undoubtedly related to the plethora of hormonal changes occurring throughout this period. Some of the hormones undergoing major changes are listed in Table 18.2. Obviously, there must be synergisms and redundancy in the hormonal signalling systems, but our approaches to date to investigate these have been relatively simple. Nevertheless, somatotropin, prolactin and glucocorticoids have the clearest identified effects which are consistent with homeorhetic controls (Bauman and Elliot, 1983; Bell and Bauman, 1997; Vernon, 1998; Chilliard, 1999). The overall mechanisms for shifting nutrient partitioning and metabolism involve alterations in the set-points for physiological responses to homeostatic controls. These alterations can be reflected by changes in the sensitivity or the magnitude of the biological response. The former is reflected by a change in the effective dose to obtain a 50% response (ED50) for the homeostatic signal whereas the latter is reflected by a change in the maximum response (Rmax) to the homeostatic signal (Kahn, 1978). Specific mechanisms for the alterations of the response to homeostatic signals can include alterations in tissue receptors and binding kinetics, changes in the intracellular signal transduction systems and effects on the expression and activity of key enzymes in the biochemical pathways. The net effect is that homeorhetic controls have tissue-specific effects on both the amounts and activity of critical metabolic enzymes and the signalling proteins that regulate them. Therefore, homeorhetic adaptations allow for chronic alterations or even redirection of physiological processes while still allowing homeostatic systems to preserve constant conditions. These same general mechanisms involving alterations in the set-points for responses to homeostatic controls have been demonstrated for many biological processes in the different physiological situations listed in Table 18.1. Several of the tissues and processes in which alterations to homeostatic controls occur during lactation are listed in Table 18.3. Feed intake provides a general example, and a number of homeostatic controls for regulating feed intake have been identified (Forbes, 1996; Langhans, 1999). These homeostatic controls are obviously functioning in non-lactating and lactating animals, but in species such as the rat and cow the set points are altered during lactation. Thus, homeostatic controls of feed intake still occur but the altered set-points allow for a greater voluntary intake so that nutrient supply more adequately meets the nutrient requirement. Insulin is an especially powerful mediator of many different physiological effects, most of which serve to acutely maintain metabolic equilibrium in the face of short-term

318

D.E. Bauman

variations in nutrient supply and demand. Thus, this acute regulatory signal is a pivotal target for chronic metabolic adaptations. As illustrated by examples in Table 18.3, many tissues have specific responses to insulin which are attenuated with the onset of lactation. This includes liver (gluconeogenesis inhibition), adipose tissue (fat synthesis), skeletal muscle (glucose uptake) and whole body (glucose oxidation) (Vernon and Sasaki, 1991; Williamson and Lund, 1994; Bell and Bauman, 1997; Vernon, 1998). These adaptations during early lactation are frequently referred to as reflecting an ‘insulin resistance’. However, judicious use of this term is needed. An attenuated response is not observed for all acute regulatory functions of insulin. For example, the antilipolytic effect of insulin is greater in lactating sheep compared with non-lactating sheep (Vernon et al., 1990), and the inhibition of whole-body rates of protein degradation is enhanced during early lactation (Tesseraud et al., 1993). Thus, the changes in response to insulin are specific for certain tissues and certain biochemical processes within those tissues, rather than representing any generalized phenomenon. Overall, the physiological adaptations in the response of various processes to insulin have the net effect of enhancing hepatic production of glucose, and sparing glucose use by nonmammary tissues, consistent with the increased glucose requirement of the mammary gland. An example of the attenuated response to insulin which occurs during lactation is presented in Fig. 18.1. In this example of glucose uptake by the hindlimb, the responsiveness to insulin is substantially reduced during lactation but the sensitivity to insulin is relatively unaltered. Similar studies on glucose and acetate uptake by adipose tissue indicates the response to insulin is virtually abolished in early lactation (Burnol et al., 1986; Vernon, 1989; Vernon and Sasaki, 1991; Chilliard, 1999). A second example is

Table 18.3. A partial list of adaptations in metabolic regulation which occur during lactogenesis and early lactation in ruminants. Tissue/processes

Homeostatic control

Feed intake

Multiple controls

Adipose tissue

Insulin

Catecholamines Adenosine Skeletal muscle

Insulin Insulin(?)

Liver

Insulin

Pancreas

Insulinotropic agents

Whole animal

Insulin

Response to altered set-points

ÇAppetite and satiety set-point

ÈLipogenesis ÈUptake of preformed fatty acids ÇStimulation of lipolysis ÇInhibition of lipolysis

ÈGlucose uptake ÈProtein synthesis ÈAmino acid uptake ÇProtein degradation

ÇGluconeogenesis

ÈInsulin release

ÈGlucose oxidation ÈGlucose utilization by nonmammary tissues

Regulation of Nutrient Partitioning During Lactation

319

Fig. 18.1. Effect of insulin on the arterio-venous difference for glucose across the hindlimb of non-lactating (嘷) and lactating (●) sheep. Results represent mean ± standard error of the mean. From Vernon (1986).

the effect of ␤-adrenergics on adipose tissue rates of lipolysis (Bauman and Elliot, 1983; Vernon and Sasaki, 1991; Vernon, 1996; Chilliard, 1999). As illustrated by the elegant studies of Guesnet et al. (1987), the ability of isoproterenol, a specific ␤receptor agonist, to stimulate lipolysis is markedly altered by stage of pregnancy and early lactation (Fig. 18.2). Somatotropin (ST) is the homeorhetic control for which mechanisms have been most extensively investigated. Much of these data come from investigations of the response to exogenous ST in lactating or growing farm animals (Bauman and Vernon, 1993; Burton et al., 1994; Etherton and Bauman, 1998). In the case of dairy cows, the studies convincingly demonstrate that exogenous bovine somatotropin (bST) results in an increase in milk yield of the treated animal and a series of coordinated adaptations in body tissues to support the greater use of nutrients for milk synthesis. Many of the coordinated adaptations are manifested by alterations in tissue responses to homeostatic signals. In fact, the changes which occur with bST treatment are similar to those illustrated in Figs 18.1 and 18.2 for the onset of lactation. While our understanding of the signal transduction systems regulating enzyme activity and gene expression is still incomplete, the specific effects have been reviewed recently for both the onset of lactation (Vernon, 1998; Chilliard, 1999) and treatment with exogenous ST (Bauman and Vernon, 1993; Etherton and Bauman, 1998). Overall, the changes which occur with the onset of lactation or the initiation of bST treatment allow for a chronic alteration of nutrient utilization. This is illustrated by the above examples. When a meal is consumed and circulatory insulin increases, less nutrients are directed to body fat reserves and other non-mammary tissues because of their altered response to insulin, and more nutrients are taken up by the mammary gland consistent with the increased milk synthesis. Likewise, if nutrient supply is

320

D.E. Bauman

Fig. 18.2. Effect of ␤-adrenergic stimulation on rates of lipolysis in sheep adipocytes during pregnancy and lactation. Isoproterenol is a specific ␤-receptor agonist and results represent mean ± standard error of a mean for different days of pregnancy (P) and lactation (L). From Guesnet et al. (1987).

inadequate the coordinated responses require a greater mobilization of energy reserves to meet the requirements associated with the increased milk synthesis, and this is accommodated by a greater response to signals which stimulate lipolysis. Thus, adaptations in the response to homeostatic signals affect metabolic processes in an orchestrated manner to match the mammary gland need for nutrients for milk synthesis.

Coordination of biological processes Coordination represents a key feature of biological regulation. It was emphasized in the original ideas which led to the concept of homeostasis (Bernard, 1878; Cannon, 1929, 1932) and was central in the more recent crystallization of homeorhesis (Bauman and Currie, 1980) and analogous concepts (Chilliard, 1986; Mrosovsky, 1990; Kuenzel et al., 1999). However, coordination of tissue processes has not been universally accepted, particularly in the early years. An alternative concept articulated

Regulation of Nutrient Partitioning During Lactation

321

by Hammond (1944) was that regulation of nutrient use represented a ‘competition’ between organs. Applying this to the onset of lactation, Hammond (1952) concluded that regulation increased the metabolic rate of the mammary gland which allowed it to ‘compete more successfully with other tissues of the body for nutriments in the bloodstream’. The concept that regulation involved competition between organs also dominated many of the early experimental approaches and the interpretation of results. One example is the early studies with bST; based on the proposed mechanism of action it was anticipated that bST treatment would create a competition between organs which would cause metabolic problems and burnout (see review by Bauman, 1999). Instead, the research demonstrated that bST treatment of dairy cows resulted in a series of coordinated responses as discussed previously, and today it is used commercially in many countries (Bauman, 1999). Recent analysis of US dairy herds (n > 200,000 lactations) demonstrated use of bST improved lactational yield and persistency consistently over the 4-year period of commercial use and animal well-being was maintained as indicated by performance, milk quality, stayability and herd-life (Bauman et al., 1999). Tepperman and Tepperman (1970) were among the first to elaborate specific examples of coordination relative to metabolic processes. They referred to this as the ‘Sherrington metaphor’ based on Sir Charles Sherrington’s (Sherrington, 1947) famous ‘principle of reciprocal inhibition of antagonistic muscles’ which stated: where two muscles would antagonize each other’s action the reflex arc, instead of merely activating one of the two, when it activates the one causes depression of the activity of the other. The latter is an inhibitory effect.

Examples of the Sherrington metaphor provided by Tepperman and Tepperman (1970) included the coordination between metabolic pathways within a cell (e.g. lipogenesis and gluconeogenesis in the rat hepatocyte) and between organs in different nutritional states (e.g. nutrient flow during fasting and re-feeding). During lactation, the exquisite nature of the coordination of biological processes is illustrated by many examples of the Sherrington metaphor. Coordination occurs among biochemical pathways within a cell as indicated by the reduction in the pathways of fatty acid synthesis and the increased importance of the pathways of lipolysis which occurs in adipose tissue with the onset of lactation (Table 18.2). The orchestrated changes are also tissue specific. For example, lipoprotein lipase activity and the uptake of preformed fatty acids as well as de novo lipogenesis and activities of key enzymes in fatty acid synthesis are reduced in adipose tissue with the onset of lactation, while these exact same enzymes and biochemical pathways are dramatically increased in the mammary gland during the same time interval. Furthermore, the coordinated responses frequently involve many different organs and physiological processes as illustrated by the aforementioned example of glucose. The substantial mammary demand for glucose which occurs with the onset of lactation is provided for by a series of orchestrated adaptations which include increased hepatic synthesis of glucose, reduced glucose uptake by several non-mammary tissues and an overall reduction in wholebody oxidation of glucose (Table 18.2). Of special importance, if the coordinated regulation of these processes is inadequate so that glucose supply fails to meet the overall needs during lactation, then a chain of metabolic events occurs leading to a metabolic disease, ketosis, and a compromise in animal well-being.

322

D.E. Bauman

The fact that biological regulation involves a series of orchestrated responses is frequently overlooked in articles relating to animal welfare. At regular intervals over the last 50 years some have expressed concern that practices to improve the productive efficiency of dairy cows may be pushing them too far, thereby compromising animal health and shortening the lifespan. Hammond (1952) and Bauman et al. (1985) reviewed the physiological limits to production, and they did not share that concern. Indeed, milk yield and productive efficiency have continued to increase throughout this century and today we have herds with annual milk yields over 15,000 kg per cow and individual cows that have produced over 27,000 kg year⫺1 (Bauman, 1999). Nevertheless, some continue to assume that genetic selection and improved management practices which lead to more efficient, high-producing animals are at variance with the physiological controls for animal well-being (Rauw et al., 1998; Broom, 1999). For example, a recent review on welfare of dairy cattle suggested that it ‘may well be necessary to stop using genetic selection and some feeding methods to increase milk yield’ because these practices have resulted in stressed cows in which ‘their normal biological functioning controls are overtaxed’ (Broom, 1999). This viewpoint clearly fails to appreciate that genetic selection and management improvements are successful because they have altered the biological controls in a coordinated manner. Rather than the biological controls being at discord with increased performance, it is the improvements in the biological control systems which are responsible for the increases in milk yield and the gains in productive efficiency. The role of coordinated responses in metabolic regulation is also important in the development of dynamic models of metabolism. Recent studies on the regulation of milk protein synthesis provide an example of this coordination. Most investigations of milk protein production in dairy cows have examined adequacy of amino acid (AA) supply. In general, results indicate that supplemental amino acids increase milk protein content and yield when the AA supply from microbial and bypass protein is inadequate, but little or no response occurs in well-fed cows (Sutton, 1989; Rulquin et al., 1995). We were interested in the endocrine regulation of milk protein synthesis, and focused on the chronic effects of insulin because of data summaries showing a high correlation between dietary energy intake and milk protein content (Sporndly, 1989). By using the hyperinsulinaemic-euglycaemic clamp technique we were able to examine the role of insulin without the confounding effects of hypoglycaemia. This technique involves intravenous infusion of insulin to achieve a constant elevated concentration and simultaneous infusion of sufficient glucose to maintain normal blood concentrations. Results indicated that a 4-day insulin clamp resulted in a dramatic increase in milk protein content and yield; in the well-fed cow milk protein yield increased by over 25% (Griinari et al., 1997a; Mackle et al., 1999). Plasma urea nitrogen concentrations were markedly reduced during the insulin clamp, demonstrating that one component of the coordinated response was a reduction in whole-body oxidation of AA (McGuire et al., 1995; Griinari et al., 1997a; Mackle et al., 1999). The means by which the mammary gland obtained sufficient AA to support the increased output of milk protein was of special interest because plasma concentrations of essential AA were reduced by 30–50% during the insulin clamp. We found that during the insulin clamp both blood flow (ml ml⫺1 of milk) and extraction of essential AA (per cent) were increased (Mackle et al., 2000). Thus, the mammary gland component of the coordinated responses included alterations in the local control of blood flow and transport mechanisms for

Regulation of Nutrient Partitioning During Lactation

323

essential AA as well as an increased ability to synthesize milk proteins. Overall, the orchestrated series of changes which occurred during the chronic insulin clamp involved adaptations in whole-body processes to allow a greater proportion of AA to be used for milk synthesis and adaptations by the mammary gland to increase AA uptake and rates of milk protein synthesis. The coordination of metabolic regulation also needs to be considered in evaluating research results. An example is research relating to the low-fat milk syndrome in dairy cows (Davis and Brown, 1970). This syndrome is characterized by a marked reduction in both yield and percentage of milk fat, and diets which cause milk fat depression (MFD) typically result in a more positive energy balance, an increase in circulating insulin, and an increase in body fat accretion. Variations in circulating insulin have no acute effects on mammary lipid metabolism of dairy cows, but as discussed earlier they can affect adipose tissue rates of lipogenesis and lipolysis. One theory for the cause of MFD is the glucogenic-insulin theory which postulates that when diets cause an increase in circulating insulin, the mammary gland is deprived of milk fat precursors due to vigorous competition by adipose tissue (see discussion by Griinari et al., 1997b). We have evaluated this theory using a 4-day hyperinsulinaemic-euglycaemic clamp described previously. Despite the substantial challenge to the mammary gland supply of lipogenic precursors imposed by the fourfold increase in circulating insulin, body metabolism was coordinated so that the rate of milk fat synthesis was relatively constant during the insulin clamp (McGuire et al., 1995; Griinari et al., 1997b; Mackle et al., 1999). Thus, our work demonstrated the coordinated regulation of nutrient partitioning, but it provided no support for the glucogenic-insulin theory. What then is the cause of dietary-induced MFD? Another theory proposes that MFD is caused by a direct inhibition of mammary fatty acid synthesis by products from incomplete or unusual biohydrogenation of polyunsaturated fatty acids in the rumen (Davis and Brown, 1970). Trans fatty acids have received special attention as the cause (Davis and Brown, 1970; Erdman, 1996), and more recently we have expanded this to include trans fatty acids and related metabolites (Griinari et al., 1998). Consistent with the trans fatty acid theory, MFD is observed when partially hydrogenated vegetable oils are abomasally infused and there is a close relationship between the decrease in milk fat percentage and the increase in milk fat content of trans-C18:1 over a wide range of diets (Erdman, 1996; Griinari et al., 1998). The typical ruminal biohydrogenation of linoleic acid to stearic acid produces cis-9, trans-11 conjugated linoleic acid (CLA) and trans-11 C18:1 as intermediates. We found that diets which cause MFD resulted in a shift in ruminal biohydrogenation with the reduction in milk fat percentage corresponding to increases in the milk fat content of trans-10 C18:1 and trans-10, cis-12 CLA (Griinari et al., 1998, 1999). To directly examine the role of the CLA on milk fat synthesis in dairy cows we abomasally infused CLA isomers. The cis9, trans-11 CLA isomer had no impact on milk fat whereas a 4-day abomasal infusion of less than 10 g day⫺1 of trans-10, cis-12 CLA resulted in over a 40% reduction in milk fat content and yield (Fig. 18.3; Baumgard et al., 2000). Furthermore, the shifts in milk fatty acid composition found with trans-10, cis-12 CLA paralleled those observed with classical MFD. Thus, our results are consistent with diet-induced MFD involving a direct inhibition of milk fat synthesis by intermediates formed in rumen biohydrogenation of polyunsaturated fatty acids. In this scenario the increase in body fat accretion which is particularly evident with MFD induced by high grain–low fibre

324

D.E. Bauman

Fig. 18.3. Temporal pattern of milk fat content during abomasal infusion of conjugated linoleic acid (CLA) isomers. Infusions represented 10 g day⫺1 of cis-9, trans-11 CLA or trans-10, cis-12 CLA. From Baumgard et al. (2000).

diets would represent a shift in nutrient partitioning which occurs as a consequence rather than a cause of the reduced mammary fat synthesis. Dietary addition of CLA has also been shown to inhibit body fat accretion in several species, but the required dose appears to be 10–20-fold greater than the dietary level which achieves MFD (Baumgard et al., 2000).

Conclusions As our knowledge of biology increases we appreciate even more the remarkable system for the regulation of metabolic processes which occurs in different physiological states. This review has re-examined the concepts of homeostasis and homeorhesis in the regulation of metabolism. Lactation provides an impressive example of homeostasis and homeorhesis in action, and examples were used to provide an overview of the integrated mechanisms. Overall, the homeostatic and homeorhetic mechanisms provide a coordinated regulation of the metabolism of different organs and tissues to ensure the proper nutrient supply to the mammary gland. In crystallizing the concept of homeorhesis and the interrelationships between homeostasis and homeorhesis, we originally concluded with a quotation (Bauman and Currie, 1980). This quotation from Duclaux, an eminent French scientist, also provides an appropriate perspective for this review. Duclaux (1920) reviewed the scientific contributions of Louis Pasteur and noted that many of his ideas were incorrect. However, he pointed out ‘ … we see clearly how much a matter of indifference it is whether a theory or a doctrine is right, provided, it incites to work, and results in the discovery of new facts’.

Regulation of Nutrient Partitioning During Lactation

325

References Bauman, D.E. (1984) Regulation of nutrient partitioning. In: Gilchrist, F.M.C. and Mackie, R.I. (eds) Herbivore Nutrition in the Subtropics and Tropics. The Science Press, Craighall, South Africa, pp. 505–524. Bauman, D.E. (1999) Bovine somatotropin: from basic science to commercial application. Domestic Animal Endocrinology 17, 101–116. Bauman, D.E. and Currie, W.B. (1980) Partitioning of nutrients during pregnancy and lactation: a review of mechanisms involving homeostasis and homeorhesis. Journal of Dairy Science 63, 1514–1529. Bauman, D.E. and Elliot, J.M. (1983) Control of nutrient partitioning in lactating ruminants. In: Mepham, T.B. (ed.) Biochemistry of Lactation. Elsevier Science Publishers B.V., Amsterdam, The Netherlands, pp. 437–468. Bauman, D.E. and Vernon, R.G. (1993) Effects of exogenous bovine somatotropin on lactation. Annual Review of Nutrition 13, 437–461. Bauman, D.E., Eisemann, J.H. and Currie, W.B. (1982) Hormonal effects on partitioning of nutrients for tissue growth: role of growth hormone and prolactin. Federation Proceedings 41, 2538–2544. Bauman, D.E., McCutcheon, S.N., Steinhour, W.D., Eppard, P.J. and Sechen, S.J. (1985) Sources of variation and prospects for improvement of productive efficiency in the dairy cow: a review. Journal of Animal Science 60, 583–592. Bauman, D.E., Everett, R.W., Weiland, W.H. and Collier, R.J. (1999) Production responses to bovine somatotropin in northeast dairy herds. Journal of Dairy Science 82, 2564–2573. Baumgard, L.H., Corl, B.A., Dwyer, D.A., Sæbø, A. and Bauman, D.E. (2000) Identification of the conjugated linoleic acid isomer which inhibits milk fat synthesis. American Journal of Physiology 278, R179–R184. Bell, A.W. and Bauman, D.E. (1997) Adaptations of glucose metabolism during pregnancy and lactation. Journal of Mammary Gland Biology and Neoplasia 2, 265–278. Bernard, C. (1878) Leçons sur les phénomènes de la vie communs aux animaux et aux végétaux. Librairie J-B. Baillière et fils, Paris, France. Blaxter, K. (1989) Energy Metabolism in Animals and Man. Cambridge University Press, Cambridge, UK. Broom, D.M. (1999) The welfare of dairy cattle. In: Aagaard, K. (ed.) Future Milk Farming. FILIDF 25th International Dairy Congress. Danish Natl. Com. IDF, Aarhus, Denmark, pp. 32–39. Brown, R.E. (1969) The conversion of nutrients into milk. In: Swan, H. and Lewis, D. (eds) The University of Nottingham Third Nutrition Conference for Feed Manufacturers. J&A Churchill Ltd., London, UK, pp. 23–29. Burnol, A.-F., Guerre-Millo, M., Lavau, M. and Girard, J. (1986) Effect of lactation on insulin sensitivity of glucose metabolism in rat adipocytes. Federation of European Biochemical Societies 194, 292–296. Burton, J.L., McBride, B.W., Block, E., Glimm, D.R. and Kennelly, J.J. (1994) A review of bovine growth hormone. Canadian Journal of Animal Science 74, 167–201. Cannon, W.B. (1929) Organization for physiological homeostasis. Physiological Reviews IX, 399–431. Cannon, W.B. (1932) The Wisdom of the Body. W.W. Norton & Company, Inc., New York. Chilliard, Y. (1986) Revue bibliographique: Variations quantitatives et métabolisme des lipides dans les tissus adipeux et le foie au cours du cycle gestation-lactation. 1re partie: chez la ratte. Reproduction, Nutrition, Dévelopment 26, 1057–1103. Chilliard, Y. (1987) Revue bibliographique: Variations quantitatives et métabolisme des lipides dans les tissus adipeux et le foie au cours du cycle gestation-lactation. 2e partie: chez la brebis et la vache. Reproduction, Nutrition, Dévelopment 27, 327–398.

326

D.E. Bauman Chilliard, Y. (1999) Metabolic adaptations and nutrient partitioning in the lactating animal. In: Martinet, J., Houdebine, L.M. and Head, H.H. (eds) Biology of Lactation. INRA Editions, Paris, France, pp. 503–552. Costa, D.P., Le Boeuf, B.J., Huntley, A.C. and Ortiz, C.L. (1986) The energetics of lactation in the Northern elephant seal, Mirounga angustirostris. Journal of The Zoological Society of London 209, 21–33. Cripps, A.W. and Williams, V.J. (1975) The effect of pregnancy and lactation on food intake, gastrointestinal anatomy and the absorptive capacity of the small intestine in the albino rat. British Journal of Nutrition 33, 17–32. Davis, C.L. and Brown, R.E. (1970) Low-fat Milk Syndrome. In: Phillipson, A.T. (ed.) Physiology of Digestion and Metabolism in the Ruminant. Oriel Press Limited, Newcastle upon Tyne, UK, pp. 545–565. Dilman, V.M. (1982) Endocrinology and diseases of aging. In: Essman, W.B. (ed.) Regulatory Processes in Clinical Endocrinology. Spectrum Publications, Inc., New York, pp. 123–161. Duclaux, E. (1920) Pasteur: The History of a Mind. (Translated by Smith, E.F. and Hedtes, F.) W.B. Saunders Co., Philadelphia. Erdman, R. (1996) Milk fat depression: some new insights. Tri-State Dairy Nutrition Conference, Fort Wayne, Indiana, pp. 1–16. Etherton, T.D. and Bauman, D.E. (1998) Biology of somatotropin in growth and lactation of domestic animals. Physiology Reviews 78, 745–761. Forbes, J.M. (1996) Integration of regulatory signals controlling forage intake in ruminants. Journal of Animal Science 74, 3029–3035. Griinari, J.M., McGuire, M.A., Dwyer, D.A., Bauman, D.E., Barbano, D.M. and House, W.A. (1997a) The role of insulin in the regulation of milk protein synthesis in dairy cows. Journal of Dairy Science 80, 2361–2371. Griinari, J.M., McGuire, M.A., Dwyer, D.A., Bauman, D.E. and Palmquist, D.L. (1997b) Role of insulin in the regulation of milk fat synthesis in dairy cows. Journal of Dairy Science 80, 1076–1084. Griinari, J.M., Dwyer, D.A., McGuire, M.A., Bauman, D.E., Palmquist, D.L. and Nurmela, K.V.V. (1998) Trans-octadecenoic acids and milk fat depression in lactating dairy cows. Journal of Dairy Science 81, 1251–1261. Griinari, J.M., Nurmela, K., Dwyer, D.A., Barbano, D.M. and Bauman, D.E. (1999) Variation of milk fat concentration of conjugated linoleic acid and milk fat percentage is associated with a change in ruminal biohydrogenation. Journal of Animal Science 77(Suppl. 1), 117–118 (abstract). Guesnet, P., Massoud, M. and Demarne, Y. (1987) Effects of pregnancy and lactation on lipolysis of ewe adipocytes induced by ␤-adrenergic stimulation. Molecular and Cellular Endocrinology 50, 177–181. Hammond, J. (1944) Physiological factors affecting birth weight. Proceedings of the Nutrition Society 2, 8–12. Hammond, J. (1952) Physiological limits to intensive production in animals. British Agricultural Bulletin 4, 222–224. Kahn, C.R. (1978) Insulin resistance, insulin insensitivity, and insulin unresponsiveness: A necessary distinction. Metabolism 27, 1893–1902. Kennedy, G.C. (1967) Ontogeny of mechanisms controlling food and water intake. In: Code, C.F. (ed.) Handbook of Physiology. Section 6: Alimentary Canal . Control of Food and Water Intake, Vol. I. American Physiological Society, Washington DC, pp. 337–351. Kuenzel, W.J., Beck, M.M. and Teruyama, R. (1999) Neural sites and pathways regulating food intake in birds: A comparative analysis to mammalian systems. Journal of Experimental Zoology 283, 348–364. Langhans, W. (1999) Appetite regulation. In: Lobley, G.E., White, A. and MacRae, J.C. (eds)

Regulation of Nutrient Partitioning During Lactation

327

Protein Metabolism and Nutrition. EEAP Publication No. 96, Wageningen Pers, Wageningen The Netherlands, pp. 225–251. Lerner, I.M. (1954) Genetic Homeostais. Oliver and Boyd, London, UK. Lewontin, R.C. (1956) Studies on homeostasis and heterozygosity I. General considerations. Abdominal bristle number in second chromosome homozygotes of Drosophila melanogaster. The American Naturalist 90, 237–255. Linzell, J.L. (1967) The magnitude and mechanisms of the uptake of milk precursors by the mammary gland. Proceedings of the Nutrition Society 27, 44–52. Lockyer, C. (1981) Growth and energy budgets of large baleen whales from the southern hemisphere. In: Mammals in the Sea. Fisheries Series. No. 5, Vol. 3, Food and Agriculture Organization. United Nations, Rome, Italy, pp. 379–487. McGuire, M.A., Griinari, J.M., Dwyer, D.A. and Bauman, D.E. (1995) Role of insulin in the regulation of mammary synthesis of fat and protein. Journal of Dairy Science 78, 816–824. Mackle, T.R., Dwyer, D.A., Ingvartsen, K.L., Chouinard, P.Y., Lynch, J.M., Barbano, D.M. and Bauman, D.E. (1999) Effects of insulin and amino acids on milk protein concentration and yield from dairy cows. Journal of Dairy Science 82, 1512–1524. Mackle, T.R., Dwyer, D.A., Ingvartsen, K.L., Chouinard, P.Y., Ross, D.A. and Bauman, D.E. (2000) Effects of insulin and post ruminal supply of protein on use of amino acids by the mammary gland for milk protein synthesis. Journal of Dairy Science 83, 93–105. McNamara, J.P. (1995) Role and regulation of metabolism in adipose tissue during lactation. Nutritional Biochemistry 6, 120–129. Millar, J.S. (1979) Energetics of lacation in Peromyscus maniculatus. Canadian Journal of Zoology 57, 1015–1019. Mrosovsky, N. (1990) Rheostasis: The Physiology of Change. Oxford University Press, Inc., New York. Nicolaïdis, S. (1983) Physiologie du comportement alimentaire. In: Meyer, P. (ed.) Physiologie Humaine. Flammarion Médecine-Sciences, Paris, France, pp. 1059–1076. Oftedal, O.T. (1993) The adaptation of milk secretion to the constraints of fasting in bears, seals, and baleen whales. Journal of Dairy Science 76, 3234–3246. Rauw, W.M., Kanis, E., Noordhuizen-Stassen, E.N. and Groomers, F.J. (1998) Undesirable side effects of selection for high production efficiency in farm animals: a review. Livestock Production Science 56, 15–33. Riedman, M. (1990) The Pinnipeds: Seals, Sea Lions and Walruses. University of California Press, Berkeley, California. Rulquin, H., Verite, R., Guinard, G. and Pisulewski, P.M. (1995) Dairy cows’ requirement for amino acids. In: Ivan, M. (ed.) Animal Science Research and Development. Canadian Society of Animal Science, Ottawa, Ontario, Canada, pp. 143–160. Sherrington, C. (1947) The Integrative Action of the Nervous System, 2nd edn. Yale University Press, New Haven, Connecticut. Sporndly, E. (1989) Effects of diet on milk composition and yield of dairy cows with special emphasis on milk protein content. Swedish Journal of Agricultural Research 19, 99–106. Sutton, J.D. (1989) Altering milk composition by feeding. Journal of Dairy Science 72, 2801–2814. Tepperman, J. and Tepperman, H.M. (1970) Gluconeogenesis, lipogenesis and the Sherringtonian metaphor. Federation Proceedings 29, 1284–1293. Tesseraud, S., Grizard, J., Debras, E., Papet, I., Bonnet, Y., Bayle, G. and Champredon, C. (1993) Leucine metabolism in lactating and dry goats: effect of insulin and substrate availability. American Journal of Physiology 265, E402–E413. Vernon, R.G. (1986) The response of tissues to hormones and the partition of nutrients during lactation. In: Research Reviews, Hannah Yearbook. Hannah Research Institute, Ayr, UK, pp. 115–121.

328

D.E. Bauman Vernon, R.G. (1989) Endocrine control of metabolic adaptation during lactation. Proceeedings of the Nutrition Society 48, 23–32. Vernon, R.G. (1996) Control of lipid mobilisation in ruminant adipose tissue. Journal of Reproduction and Development 42, 118–122. Vernon, R.G. (1998) Homeorhesis. In: Research Reviews, Hannah Yearbook. Hannah Research Institute, Ayr, UK, pp. 64–73. Vernon, R.G. and Sasaki, S. (1991) Control of responsiveness of tissues to hormones. In: Tsuda, T., Sasaki, Y. and Kawashima, R. (eds) Physiological Aspects of Digestion and Metabolism in Ruminants. Proceedings of the 7th International Symposium on Ruminant Physiology. Academic Press, San Diego, California, pp. 155–182. Vernon, R.G., Faulkner, A., Hay, W.W., Jr, Calvert, D.T. and Flint, D.J. (1990) Insulin resistance of hind-limb tissues in vivo in lactating sheep. Biochemistry Journal 270, 783–786. Waddington, C.H. (1942) Canalization of development and the inheritance of acquired characters. Nature 150, 563–565. Waddington, C.H. (1953) The ‘Baldwin effect,’ ‘genetic assimilation’ and ‘homeostasis.’ Evolution 7, 386–387. Waddington, C.H. (1957) The Strategy of the Genes. The MacMillan Company, New York. Wade, G.N. and Schneider, J.E. (1992) Metabolic fuels and reproduction in female mammals. Neuroscience and Biobehavioral Reviews 16, 235–272. Welch, G.R. (1991) Thermodynamics and living systems: Problems and paradigms. Journal of Nutrition 121, 1902–1906. Williamson, D.H. (1980) Integration of metabolism in tissues of the lactating rat. FEBS Letters 117, K93–K105. Williamson, D.H. and Lund, P. (1994) Cellular mechanisms for the regulation of adipose tissue lipid metabolism in pregnancy and lactation. In: Allen, L., King, J. and Lönnerdal, B. (eds) Nutrient Regulation During Pregnancy, Lactation and Infant Growth. Plenum Press, New York, pp. 45–70. Williamson, D.H., Lund, P. and Evans, R.D. (1995) Substrate selection and oxygen uptake by the lactating mammary gland. Proceedings of the Nutrition Society 54, 165–175.

19

The Insulin-like Growth Factor (IGF) System in the Mammary Gland: Role of IGFBP-3 Binding Protein

C.R. BAUMRUCKER Department of Dairy and Animal Science, Penn State University, Pennsylvania, USA

Introduction to the insulin-like growth factor (IGF) system The IGF system (Fig. 19.1) consists of the three ligands, insulin, IGF-I and IGF-II, three corresponding high affinity receptors, and six IGF binding proteins (IGFBP) that associate with the IGFs. The IGF ligands are known to be ubiquitous and act as mediators of growth, development and differentiation (Daughaday and Rotwein, 1989; LeRoith et al. 1995). The three peptides have around 50% amino acid homology (LeRoith, 1991). Insulin, synthesized in pancreatic ␤-cells as proinsulin, is degraded to insulin and the C-peptide, both of which are found in the circulation. IGFs, which are synthesized in many tissues, retain the C-peptide and have an extended carboxy terminus sometimes called the D-peptide (Daughaday and Rotwein, 1989). While insulin circulates in the blood freely and has a short half-life, IGFs circulate at higher concentrations, but are largely bound to one of six IGFBPs. The potential for the IGFs to bind to any one of the six known high affinity IGFBPs imparts potential for signal inhibition, signal potentiation, extended ligand half-life and specific tissue targeting (Zapf, 1995).

IGF receptors The biological effects of the IGFs are mediated through specific cell surface protein receptors. These include the IGF-I receptor (IGF-IR), the insulin receptor, and the IGF-II (IGF-IIR)/mannose-6-phosphate receptor (m-6-pR). The IGF-IR is structurally similar to the insulin receptor. Both couple tyrosine kinase activity to a series of intracellular signalling pathways (LeRoith et al., 1993). The similarity between these ligands and receptors explains much of the ambiguity of research findings when high levels of either ligand are applied to an experimental system. Under such conditions, IGF-I and insulin are known to cross-react with the other species’ receptor and trigger intracellular signal cascades associated with both receptors. In addition, both IGF ligands © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

329

330

C.R. Baumrucker

Protease IGFBP-3

IGFBP-1

IGFBP-2

IGFBP-5

IGFBP-4

IGFBP-6

IGF binding proteins IGFBP-rP(1-4)

Ligand

Receptor

IGF-II

Type 2 IGF-R

IGF-I

Insulin

Type 1 IGF-R

IGFBP-3

Insulin-R

BP-3BP

Function

Unknown

Growth

Metabolism

Apoptosis

Fig. 19.1. Figure showing the components of the insulin-like growth factor (IGF) system. Arrows indicate the binding and interactions that occur between the components.

bind with high affinity with the IGF-IR and IGF-I has a reduced affinity with the IGF-IIR. The IGF-IIR is not linked to tyrosine kinase activity, but rather appears to be coupled to a G protein (Nishimoto, 1993). The receptor expression is developmentally regulated, with high expression in fetal and neonatal tissues (Nissley et al., 1993). The IGF-IIR is also the cation-independent m-6-pR, although the ligand binding sites are separate. The latter is known to function as a lysosomal enzyme targeting protein, the activation of transforming growth factor-␤ (TGF-␤) (Dennis and Rifkin, 1991), and the degradation of IGF-II (Oka et al., 1985). Most recently, it has been reported that the m-6-p/IGF-IIR is also a receptor for retinoic acid (Kang et al., 1997).

IGFBPs The majority of IGF ligands are bound to IGFBPs in vivo. Unlike the transmembrane IGF receptors, the IGFBPs are secreted. They are present in blood serum, all biological fluids, and conditioned media from all in vitro cell cultures. While IGF-I and IGF-II bind to all IGFBPs with high affinity (~10⫺10 M), insulin does not bind to IGFBPs. Six distinct human IGFBPs have been identified and have been termed IGFBP-1 through to IGFBP-6 (Anonymous, 1992). Additionally, four IGFBP-related proteins that exhibit reduced binding affinity (four to tenfold less) for IGFs have been recently identified (Baxter et al., 1998). All of these proteins exhibit the presence of the specific IGFBP motif (GCGCCXXC) (Rosenfeld, 1998). Table 19.1 shows the proposed nomenclature for the superfamily of IGFBPs.

The Insulin-like Growth Factor (IGF) System in the Mammary Gland

331

Table 19.1. Insulin-like growth factor binding protein (IGFBP) superfamily. Name

IGF affinity

Other/previous names

IGFBP-1 IGFBP-2 IGFBP-3 IGFBP-4 IGFBP-5 IGFBP-6

High High High High High High

PP12, BP-28

IGFBP-related proteins (CNN family) IGFBP-rP1 IGFBP-rP2 IGFBP-rP3 IGFBP-rP4

Low Low Low Low

Mac25, TAF, PSF, IGFBP-7 CTGF, fisp-12, IGFBP-8 Nov, IGFBP-9 Cry-61, cyr10, IGFBP-10

BP-53

The role of IGFBPs in the regulation of the IGF system has been the source of many reviews (Jones and Clemmons, 1995; Kelley et al., 1996; Murphy, 1998) and concepts relating to cancer have emerged (Werner and LeRoith, 1996). Generally, IGFBP bio-action is thought to be: (i) modulation action of IGF-I or IGF-II by competition with the IGF-IR or IGF-IIR; (ii) alteration of the action of IGF-I or IGF-II; (iii) transport and/or extended half-life of IGFs; (iv) interactions with other growth factor systems or (v) IGF-independent action via (a) nuclear localization sequence, (b) IGFBP binding proteins, or (c) yet unknown mechanisms (Murphy, 1998). The most important insight into these suggested bio actions is that they mainly emerge from in vitro cell culture experiments. The IGFBP related proteins (IGFBPrP1–4) are a family of proteins (CCN family; connective tissue growth factor; cef10/cry61 and nov) (Bork, 1998) that have been recently identified as low-affinity IGFBPs. These proteins are expressed in cells within minutes (called immediate-early genes) of cellular stimulation by growth factors or transforming oncogenes (Williams et al., 1992). Because of the high level of sequence similarity among the members of the CCN family, they probably have common molecular functions. Because of their low IGFBP affinity and rather new entry into the IGF arena, these proteins will not be considered in the remainder of this chapter. The differential nature of responses to IGF ligands in various experiments in different tissues has amplified the interest of the IGFBPs in these interactions. Because IGFBP bind to IGFs with high affinity that effectively competes with the IGF-IR (Clemmons, 1997), they may be viewed as agents that prevent IGFs action. Apparently to counteract this blocking action, proteases have been shown to attack IGFBPs and release free IGF resulting from the altered IGFBP affinity (Fowlkes, 1997). For example, IGFBP-5 that is released by cells in culture is rapidly cleaved into 23 and 16 kDa fragments by a cation-dependent serine protease (Imai et al., 1997). It is not known if this protease is present in mammary cell cultures. Other means of altering IGFBP action with IGFs is by protein phosphorylation (IGFBP-1) (Jones et al., 1991) and cell surface binding that lowers IGFBP-1 and IGFBP-3 affinity (Mccusker et al., 1990).

332

C.R. Baumrucker

While IGFBP-1 binds to integrin via an arginine–glycine–aspartate (RGD) sequence, the binding of IGFBP-3 to the cell surface has only recently been shown to be due to specific binding to a number of low molecular weight proteins (Oh and Rosenfeld, 1998). Although the ability of IGFBPs to bind IGFs is thought to be an inhibitory action, IGF action has been potentiated with low concentrations of IGFBP-3 and is thought to occur by concentrating the IGFs on the cell surface in the region of the IGF receptors (Rechler, 1993). Because the affinity of IGFBP-3 is reduced tenfold when bound to the cell surface, the current concept for IGF potentiation suggests that endogenous IGF-I interaction with the bound IGFBP-3 favours a slow release and possibly avoiding down-regulation or the type-1 receptors.

IGFBP-3 Because IGFBP-3 is the major IGFBP and a reliable recombinant source became available some time ago, innumerable studies have focused upon its biological action. Although the binding and inhibition of IGF action can explain the inhibition of cellular growth by the application of exogenous IGFBP-3, several lines of evidence have emerged to suggest that other mechanisms may also be involved in the actions of the IGF system. First, IGFBP-3 has been shown to negatively regulate cell proliferation through an IGF receptor-independent pathway. Stable transfection of BALB/c 3T3 mouse fibroblasts with human IGFBP-3 cDNA decreased the rate of cell proliferation (Lamson et al., 1993). Studies using a fibroblast cell line developed from mice with targeted disruption of the IGF-I receptor established that the growth inhibitory effects of IGFBP-3 do not involve IGF binding to the IGF-IR (Valentinis et al., 1996). IGF receptor-independent actions of IGFBP-3 have also been observed in breast cancer cells (Oh et al., 1993b) and are believed to be responsible for the predisposition of breast cancer cells to apoptosis (Gill et al., 1997). Secondly, both IGF-I and IGFBP-3 have been shown to be present in the cell nucleus (Radulescu, 1994; Li et al., 1998; Peralta Soler et al., 1990; Radulescu and Wendtner, 1993) including that of breast cancer cells (Schedlich et al., 1998) and primary cultures of mammary cells (Baumrucker et al., 1999). In one of these reports (Li et al., 1998), IGF-I and IGFBP-3 have been observed to be co-localized in the nucleus of opossum kidney cells. Based on the presence of a putative nuclear localization sequence in the structure of IGFBP-3, it was proposed that IGFBP-3 might act as a carrier for IGF-I (Li et al., 1998). A recent report indicates that IGFBP-3 competes for binding with the TGF-␤ receptor (Leal et al., 1997) and the retinoic acid receptor (Kang et al., 1997). Utilizing a yeast two-hybrid system, Murphy (1998) suggests that a cDNA encoding latent TGF-␤-binding protein 1 interacts with IGFBP-3. In support of the IGFBP-3 role associated with negative growth regulation action, the expression of IGFBP-3 has also been shown to be regulated by other growth-inhibitory (and apoptosis-inducing) agents such as TGF-␤ (Gucev et al., 1996; Huynh et al., 1996; Rajah et al., 1997) antioestrogens (Buckbinder et al., 1995) and tumour necrosis factor-␣ (TNF-␣) (Yateman et al., 1993). In normal human mammary epithelial cells, growth inhibition by atRA is independent of p53 expression (Seewaldt et al., 1999).

The Insulin-like Growth Factor (IGF) System in the Mammary Gland

333

IGFBP proteases Degradation of IGFBP-3 by specific proteases has been reported in the milk and serum of humans and rats (Lamson et al., 1991). Generally, protease modified binding proteins have decreased affinity for IGFs. Plasmin, a protease found in bovine milk (Politis et al., 1988) that has been projected to be involved in mammary involution, is an IGFBP protease (Politis et al., 1995). Since the initial discovery of the IGFBP-3 protease, other proteases have been described that cleave IGFBP-2, -4, -5 and -6 (CollettSolberg and Cohen, 1996). These proteases are known as prostate-specific antigen (Cohen et al., 1994), cathepsins (Conover and De Leon, 1994), matrix metalloproteinases and others (Rajah et al., 1995). While the proteases have been suggested as a local tissue capacity to free bound IGF from the IGFBP and thereby provide free IGF to the IGF-IR, independent actions of IGFBP fragments remain a possibility since Yamanaka et al. (1997) demonstrated that an IGFBP-3 fragment derived from protelolytic attack binds to insulin.

IGF system change and mammary physiology Blood versus milk changes The initial focus of animal scientists towards the IGF system was linked to the availability of tools to conduct temporal survey analyses and the drive to explain the galactopoietic effect of bovine growth hormone (Bauman and Vernon, 1993). Both IGF-I and -II have been characterized in the ruminant blood relative to the stage of lactation. Ronge et al. (1988) showed that blood IGF-I concentration was inversely related to milk production with a drastic drop after parturition followed by a gradual increase as lactation persisted. IGF-II does not appear to be affected by lactation (Vicini et al., 1991). During the dry period of a pregnant animal, serum IGF-I is high in concentration, but declines rapidly after parturition while IGF-II is largely unchanged (Vega et al., 1991). While exogenous injections of bovine growth hormone increase IGF-I concentrations in the blood, the increase occurs in a period when normal IGF-I concentrations are very low and the low concentration was partially attributed to lower liver mRNA levels suggesting decreased synthesis (Sharma et al., 1994). However, blood clearance (tissue utilization) was not examined. IGFBP concentration in the serum of ruminant species has been reported to not change for IGFBP-3 during lactation, but is higher during the pre-partum period (Vicini et al., 1991; Sharma et al., 1994). We reported prior to these reports that blood IGFBP activity was very high during the pre-partum colostral phase and then declined to lower levels essentially paralleling the IGF-I pattern (Vega et al., 1991). Recently, we have shown that the bovine blood and milk IGFBP, identifiable by Western ligand blots, are IGFBP-2, -3, -4 and -5 (Gibson et al., 1999). The blood identification confirms that established by others (Funston et al., 1995; Roberts et al., 1997). Figure 19.2 shows a survey of circulating serum milk IGFBP occurring during the full course of lactation. As has been previously demonstrated, Fig. 19.2c shows that IGFBP-3 is the predominant IGFBP in bovine circulation (Roberts et al., 1997). The comparison between milk and blood from the same animals indicates that although IGFBP-3 and

334 C.R. Baumrucker

Fig. 19.2. Comparison of the changes in insulin-like growth factor binding proteins (IGFBP) in (a) milk colostrum, (b) milk whey, and (c) blood serum during lactation. Cows were sampled for blood and milk at 2-week intervals. Western blots utilizing [125I]IGF-II as a ligand. Data was standardized by the use of an internal blood serum standard. Lines are regressions for each IGFBP. n = 33 cows.

The Insulin-like Growth Factor (IGF) System in the Mammary Gland

335

IGFBP-2 change in both blood and milk during the course of lactation, their patterns of change are significantly different (Fig. 19.2b and c). While milk is approximately tenfold less in concentration of IGFBPs than that of blood (Fig. 19.2b versus 2c), the pattern clearly shows that blood IGFBP changes (Fig. 19.2c) are not reflected in milk changes (Fig. 19.2b). Milk IGFBPs all drop precipitously after the colostrum phase and the onset of copious milk secretion and IGFBP-3 shows an increase towards the end of the milk production period. The inverse relationship with milk production is evident. Milk IGFBP-2 is relatively lower in concentration in both fluids and may exhibit some changes, but the variation between animals precludes interpretation. Milk and blood IGFBP-4 and -5 change little during the course of lactation. While nutrition has an impact upon the IGF axis (Thissen et al., 1994), feed restriction experiments show little influence over IGFBP-3 serum concentrations (McGuire et al., 1995). Because IGF-I is low when serum bovine growth hormone concentrations are high, the growth homone/IGF axis is said to be uncoupled. The 33 animals used for the data shown in Fig. 19.2 were fed adequately relative to NRC requirements (National Research Council, 1988). Statistical analysis of the milk changes for all of the IGFBPs showed that there were two main effects upon IGFBP changes. The first was time of lactation and the second was an effect of pregnancy. The repetitive sample study with 33 cows from the Pennsylvania State University dairy herd, shown in Fig. 19.2, had eight cows that did not become pregnant. Figure 19.3a–c shows that there are large differences between the appearance of IGFBPs in milk of cows that do not become pregnant (Fig. 19.3a) compared with those that become pregnant (Fig. 19.3b). Most notable is that the cows that do not become pregnant show high concentrations and higher variances of IGFBP-3 and -2 in their milk. This is not reflected in the IGFBP concentrations occurring in blood (Fig. 19.3c and d). These findings support three major concepts. First, some event related to pregnancy is communicating to the mammary gland to alter milk IGFBP profiles; second, that milk changes occur independent of blood changes clearly suggests that the mammary gland regulates IGFBPs occurring in milk; and lastly, the blood changes, being different from that of milk, indicate that blood IGFBPs must be reflections of more than mammary gland tissue and therefore changes in blood IGFBP components will not reflect impacts upon the mammary gland. Finally, many investigations with humans and rodents have shown that IGFBPs may be affected by the presence of proteases (Giudice, 1995). These enzymes have been shown to specifically modify IGFBPs so that they exhibit lowered binding of IGFs such that Western blotting may fail to detect them. Since our recently published characterization of IGFBPs in blood and milk utilized Western blotting, we needed to know if the milk IGFBP patterns were influenced by the presence of proteases. Extensive investigations for the presence of the pregnancy-related IGFBP-3 protease reported for the rodent and human (Staley, 1998) has been negative for bovine blood and milk. Thus, the changes observed in milk IGFBP-3 are not due to protease activity. The presence of other IGFBP proteases is currently unknown.

336

C.R. Baumrucker

Fig. 19.3. Comparison of insulin-like growth factor binding proteins (IGFBP) in milk whey and blood serum of pregnant and non-pregnant cow changes during lactation. (a) Cow milk whey from non-pregnant cows and (c) serum (n = 8). (b) Cow milk whey from pregnant cows and (d) serum (n = 25). Samples are the same as those shown in Fig. 19.2. Data was standardized by the use of a internal blood serum standard. Lines are regressions for each IGFBP.

The Insulin-like Growth Factor (IGF) System in the Mammary Gland

337

Local IGF ligands The initial interest in circulating IGF system components soon expanded when information began to accumulate indicating that all tissues and cells have the capacity to synthesize and secrete a number of IGF system components. Thus, the autocrine or paracrine concept of IGF system component regulation upon mammary gland physiology was expanded. To conduct these studies, tissue, organ or cell cultures are favoured. Mammary tissue of pregnant and lactating cows shows the presence of IGF-I mRNA (Glimm et al., 1992; Sharma et al., 1994). Subsequent studies with tissue indicated that stroma cells (Yee et al., 1989), not secretory epithelial cells, were responsible for the tissue production of IGF-I. In contrast, IGF-II appears to be synthesized by rodent mammary cells (Campana et al., 1994) and normal bovine mammary cells in vitro (Baumrucker et al., 1993).

IGF receptors High-affinity receptors for IGF-I, IGF-II and insulin have been demonstrated for bovine mammary tissue and mammary epithelial cells (Oscar et al., 1986; Hadsell et al., 1990). Studies have shown that while type-1 IGF-IR binding increases with lactogenesis (Dehoff et al., 1988), this is an apparent increase that results from a decrease that occurs during the pre-partum period (Hadsell et al., 1990). Although the type-2 IGF-II receptor shows approximately fivefold greater binding (number of receptors), no microsomal binding changes were observed during lactation (Hadsell et al., 1990). Since the mammary type-2 receptor does not change during the course of lactation, its capacity to invoke changes in the mammary gland becomes less significant, yet it probably contributes to the disappearance of IGFs from the cellular environment. Perhaps this is the mechanism of IGF appearance in milk (Prosser and Fleet, 1992; Donovan et al., 1995). The significance of the type-1 receptor change (1.6-fold increase) is perhaps debatable. This is apparent when considering insulin receptor action. Characterization of the insulin receptor has shown that less than 5% of receptor binding leads to maximal insulin stimulated events (White and Kahn, 1994; Taylor et al., 1996) leading to the concept of spare receptors. Is this also true for the type-1 receptor? Studies by Neuenschwander et al. (1995) with breast cancer cells (MCF-7) indicated that there appears to be no spare receptors for the type-1 receptors, at least in MCF-7 cells, since the antisense suppression of the expression of the type-1 receptor translated directly into a linear decrease in cellular growth in the presence of a constant supply of IGF-I. Thus, the decline observed for the type-1 receptor in mammary tissue during the late pre-partum period may be a signal for less growth and perhaps provide more opportunity for cellular differentiation. This raises the question, what is the IGF signal to the mammary gland: growth, differentiation, or apoptosis? Three transgenic studies have shed some light upon the explanation of this question. When milk protein promoters were utilized to express excess IGF-I (and analogues) during lactation, no significant difference was observed in the mammary gland milk production (pup weight gain) or structural morphology (Brem et al., 1994; Benito et al., 1996; Hadsell et al., 1996) Some reports of smaller alveoli were reported. The main effect observed is the delay in cellular decline during

338

C.R. Baumrucker

involution providing strong support for a role of IGF-I in the resistance of apoptotic mechanisms, at least during involution. In light of the low IGF-I in milk and blood during early lactation, perhaps the increase in IGF-I induced by bovine somatotropin (bST) treatments are decreasing lactational apoptosis and thereby increasing milk production. In the cow, although circulating levels of IGF are increased by bovine growth hormone, the overall pattern of IGF-I concentration is inversely related to levels of milk production (Ronge et al., 1988). This correlative data may indicate that the IGFs are low during peak lactation in order to allow maximal differentiation and continued milk production from those cells established into lactation. However, systemic circulation of IGFs is not the only source of IGF for mammary tissue. Mammary stromal fibroblasts are a source of IGF-I while epithelial cells have not been shown to have this capacity (Glimm et al., 1988; Yee et al., 1989). This suggests that although IGF-I is synthesized in mammary tissue, the level of synthesis is low and discourages the concept that local IGF-I production accounts for the galactopoietic potential during lactation. On the other hand, we have demonstrated that IGF-II is synthesized and regulated in mouse mammary epithelial cells (COMMA-D1) (Campana and Baumrucker, 1994) and primary cultures of bovine mammary epithelial cells (Fig. 19.4). 25,000

IGF-II secreted (pg ml–1)

20,000

15,000

10,000

5,000

C o (5 rtis 00 o ng l/pr /2 ola In 0 ct ng in (5 su 0 lin ) ng /c /5 ort 00 is ng ol/p /2 ro 0 la In ng ct (5 sul ) in 0 in/ ng co /5 rti 00 so ng l/p /5 rol 00 ac ng tin )

SF M

co nt ro l

0

Fig. 19.4. Secretion of insulin-like growth factor II (IGF-II) into conditioned media (CM) from primary cultures of bovine mammary epithelial cells. CM was changed each day and was analysed for IGF-II by RIA after acetic acid column separation. Values are the mean ± standard error of a mean (n = 3).

The Insulin-like Growth Factor (IGF) System in the Mammary Gland

339

How can receptor numbers and ligand changes be reconciled with changes in milk production? We believe that changes in IGFs and their receptors need to be considered in the whole IGF system (Fig. 19.1). It has been demonstrated that the application of IGFs and other endocrine factors to most tissues and mammary cells (Skaar and Baumrucker, 1993; Cohick, 1998) stimulates a change in the appearance of IGFBPs.

Mammary cell IGFBP synthesis Numerous sources show that IGFBPs are synthesized and secreted by mammary epithelial cells (Clemmons et al., 1990; Figueroa and Yee, 1992; Werner and LeRoith, 1996). This is not surprising since almost all mammalian cells synthesize and secrete IGFBPs (Jones and Clemmons, 1995). IGFBP-3 and IGFBP-2 have been identified by Western ligand blot analysis in bovine mammary pre-partum secretions (McGrath et al., 1991). Recently, we have demonstrated that primary cultures of bovine mammary epithelial cells secrete IGFBP-2, -3, -4 and -5, which by molecular weight analysis, appear to be identical to those shown in bovine milk (Gibson et al., 1999). Several hormones and growth factors regulate the synthesis and secretion of IGFBPs by mammary epithelial cells. These factors that affect mammary IGFBPs (IGFs, epidermal growth factor, prolactin, cortisol, retinol) are also known to be primary effectors of mammary growth and differentiation (Fielder et al., 1992; Skaar and Baumrucker, 1993). We have utilized frozen mammary tissue and mammary epithelial cells from bovine (MEBo) frozen recovered cells for subsequent experiments described below. These cells were prepared from bovine mammary tissue by first making mammary acini (Baumrucker et al., 1988) and then freezing the cells for subsequent culture and passages. Figure 19.4 shows that while IGF-II is secreted by bovine mammary cells, lactogenic hormones (cortisol or prolactin) appear to inhibit production, while insulin is a stimulator as shown for the rodent (Campana et al. 1994). Low levels of insulin (50 ng ml⫺1) do not displace IGF-I from the type-1 receptor in mammary tissue (Hadsell et al., 1990) suggesting that this change is attributable to the insulin receptor. Figure 19.5 shows that primary cultures of bovine mammary epithelial cells secrete the four IGFBPs into conditioned media over time and that IGFBP-3 remains the dominant IGFBP in serum-free media (Fig. 19.5b). Furthermore, we show that retinoic acid, one of the known IGFBP-3 regulatory agents, stimulates IGFBP-3 appearance with time in culture. Cohick and Turner (1998) recently utilized a SV-40 large T antigen immortalized bovine mammary cell line (MAC-T) (Huynh et al., 1991) to demonstrate some of the regulatory factors in IGFBP expression in these cells. Most interesting is that the pattern of IGFBP released by these cells in serum-free media was IGFBP-2, -4 and -6. No IGFBP-5 and little IGFBP-3 is detectable unless the latter was stimulated by IGF-I. The lack of IGFBP-3 is similar to other immortalized mammary cells such as MCF-7 cells where IGFBP-3 is also low or undetectable. However, MCF-7 cells (Shang et al., 1998, 1999) and primary bovine mammary cells (Fig. 19.5) are stimulated to produce IGFBP-3 with retinoic acid while the MAC-T cells are stimulated by IGF-I (Cohick and Turner, 1998). Thus, the regulation of IGFBP in cell cultures shows differential effects, either due to immortalization or to culture conditions.

340

C.R. Baumrucker

Fig. 19.5. Primary cultures of bovine mammary epithelial cells secrete insulin-like growth factor binding proteins (IGFBP) into conditioned media. Cells were plated on plastic in serum-free media and media changed every 2 days. IGFBPs were detected by Western blots utilizing [125I]IGF-II as a ligand. All trans retinoic acid (1 ␮M; atRA) stimulates the appearance of IGFBP-3 in the conditioned media after 12 days of treatment.

IGFBP-3 binding to membrane proteins It is now established that IGFBP-3 binds to membrane-bound proteins (Oh et al., 1993; Hodgkinson et al., 1995) Recombinant human (rh) IGFBP-3 binds to membrane proteins of 50, 26, and 20 kDa from breast cancer cells (Hwa et al., 1999). A personal communication with Youngman Oh indicated that although these proteins are detectable, the expression is low. In addition to these proteins, TGF-␤ receptor-V is also reported to be an IGFBP-3 receptor (Leal et al., 1997). A recent study (Schedlich et al., 1998) in which mutant IGFBPs were produced indicated that mutants of rhIGFBP-3 exhibit differential binding to cell surfaces and concluded that residues 228–232 were essential for cell association. Schedlich et al. (1998) also produced a mutant RGD IGFBP-3 protein (not found in the wild type) and demonstrated that

The Insulin-like Growth Factor (IGF) System in the Mammary Gland

341

not all binding was due to the introduced RGD sequence that delineates the integrin binding site (Ruoslahti and Pierschbacher, 1987). It has been speculated that the mechanism of action of IGFBP-3 may be attributed to its ability to bind the serum and autocrine IGF-II and prevent action at the type-1 IGF receptor as shown with a mouse mammary cell line (Campana et al., 1994) and the primary bovine cells shown in Fig. 19.5. However, the presence of a type-1 receptor-independent action of IGFBP-3 cannot be excluded and we now have descriptive evidence for such a mechanism as reported in preliminary data described below.

IGFBP-3: IGF-I dependent versus independent action Because IGF-I acts as a survival factor through stimulation of the type-1 receptor (Stewart and Rotwein, 1996), the induction of apoptosis by IGFBP-3 could be explained by an IGF-dependent (sequestering IGF-I) or IGF-independent mechanism. While some studies suggest the IGF-dependent pathway (Rajah et al., 1997), others show the latter (Gill et al., 1997).

Evidence of an IGFBP-3 binding protein (BP-3BP) MEBo microsomes (Pocius et al., 1984) were solubilized with SDS non-reducing loading buffer and loaded onto a SDS-PAGE gel. The gels were electroblotted to nitrocellulose and iodinated rhIGFBP-3, with and without excess unlabelled rhIGFBP-3, applied to the blots. The specific binding of 125I rhIGFBP-3 to membrane-bound microsomal proteins is shown in Fig. 19.6. Solubilized membrane proteins from cow mammary microsomes showed a number of BP-3BP that range from approximately 65 to 7 days) resting, to the blood (Adams and Cripps, 1977; Haig et al., 1989; Emery et al., 1991; S.J. McClure, unpublished observations). There are also changes in the local nerves. The autonomic innervation of the gastrointestinal tract is complex and extensive, consisting of both extrinsic and intrinsic nerves. The density of intrinsic nerve cell bodies within the gut is very high, with numbers comparable to the total number of neurons within the spinal cord. Sheep immune to Trichostrongylus colubriformis showed increases in the number and metabolic activity of enteric nerve fibres following challenge infection, suggesting that the nervous system can be primed by exposure to nematodes in a similar manner to the priming of the immune system (Stewart et al., 1995b). In addition, immune and nervous systems can synthesize and respond to shared chemical mediators. Thus it is probable that the local and central nervous systems have a role in integrating the anamnestic immunological, muscular and physiological changes that follow worm challenge into a coordinated and flexible protective response. It is not yet clear which of the many associated cellular responses are protective, and some have been depleted in vivo in an attempt to further define the protective mechanism. Depletion of CD4+ helper T-cells during challenge of sheep or goats immunized by viable infection or non-viable vaccines impaired the rejection of Haemonchus contortus, suggesting a role for CD4+ cells in protection (Gill et al., 1993; Karanu et al., 1997). Depletion of CD8+ or WC1/Tcr␥␦+ T-cells during induction of immunity to T. colubriformis both resulted in enhanced rejection of worms, suggesting that these cells may be involved in the slowness of normal induction (McClure et al., 1995). The conventional method for reducing resistance to worms is administration of glucocorticoids, but a range of cellular responses is affected, and a specific mechanism has not been identified.

Humoral responses Systemic Protective immunity in sheep against abomasal and intestinal worm species is associated with early increases in worm-specific antibody and IgA concentrations in local efferent lymph (Smith et al., 1984, 1985; S.J. McClure, unpublished observations), and with elevated levels of all isotypes of antibody in serum (McClure et al., 1992; Pfeffer et al., 1996; Shaw et al., 1998). However, with the possible exception of IgA antibody in Ostertagia infection (Stear et al., 1996), serum antibody level in sheep

428

S.J. McClure et al.

exposed to viable infection is not consistently predictive of protective immunity. It may be that serum antibody is particularly relevant for protection against the blood-sucking worms, as immunity to H. contortus can be transferred passsively with serum in sheep vaccinated with novel antigens (Smith, 1993). Rapid rejection is associated with a temporary decrease in serum antibody level coinciding with a rise in mucus levels of IgG antibody (McClure et al., 1992), presumably reflecting a protease-induced exit of blood proteins into the gut (Miller, 1996). Local Immune sheep after challenge have increased concentrations of worm-specific antibody in gut tissue and mucus, with all isotypes represented in the increase, and an increase in the number of cells with surface-bound IgE (McClure et al., 1992; Pfeffer et al., 1996). There are suggestions that IgG1 and IgE correlate best with protection against Trichostrongylus species, but more studies are required to confirm this.

Neuropeptides The role of neuropeptides in immunity to worms has not been investigated in vivo. However, the neuropeptides employed by the enteric nervous system (Substance P, somatostatin, vaso-intestinal peptide and ␤-endorphin) rendered mucosal mast cells more sensitive to limiting concentrations of worm antigens in vitro. These peptides also enhanced the in vitro proliferation to worm antigen of lymphocytes from mesenteric lymph node and prefemoral efferent lymph of immune sheep but did not affect specific antibody production by these cells (Stewart et al., 1995a, 1996).

Inflammatory mediators Local tissue concentrations and secretion into the intestine of mediators such as leucotrienes, 5-hydroxytryptamine and histamine increase during immune rejection of T. colubriformis (Steel et al., 1990; Jones et al., 1990). The administration of glucocorticoids immediately before challenge abrogates immunity to H. contortus and T. colubriformis; however, injection of more specific antagonists of leucotrienes, histamine, platelet-activating factor or phosphodiesterases failed to affect worm rejection (Adams, 1988; Jackson et al., 1988; Emery and McClure, 1995).

‘Non-specific’ mechanisms A number of local mechanisms important in the rejection of gut parasites are non-specific in effect but immunologically specific in induction. They are thus adaptive responses, developing after exposure to the antigens, and possibly only after exposure to the viable parasite. They may partly explain the observation that, to date, non-viable vaccines are less protective against browsing worms than are viable infection or irradiated larval vaccines.

Host Resistance to Gastrointestinal Parasites

429

Mucus Mucus per se has been proposed to have a protective role in trapping incoming worms and preventing their establishment (reviewed by Rothwell, 1989). Protective immunity to T. colubriformis in Merinos is associated with increased numbers of goblet cells and increased quantities of mucus in the jejunum (R.G. Windon, personal communication). The quality of the mucus also changes in sheep immune to gastrointestinal nematodes, with altered composition of the muco-polysaccharides and -proteins, the leucotriene content (Douch et al., 1983; Jones et al., 1990, 1994), and the content of various as yet unidentified molecules with inhibitory effects on nematodes. Mucus also stabilizes and prolongs the biological activity of inflammatory mediators (W.O. Jones, unpublished observations). The induction of these mechanisms is not well understood. Peristalsis Immunity in sheep to gut worms appears to be associated with increased enteric nerve fibre number and metabolic activity (Stewart et al., 1995b), and with hypertrophy and hyper-contractility of local smooth muscle (Tremain and Emery, 1994; S.J. McClure, unpublished observations). In rodents, gut smooth muscle function is subject to modulation by the immune system, with T lymphocytes, particularly those expressing CD4, implicated in the alteration of smooth muscle contractility seen by 6 days after infection with Trichinella spiralis (Vallance and Collins, 1998). These adaptive changes in gastrointestinal motility persist for some time after exposure to worms, and in addition to hustling incoming larvae, may have wider and more long-term repercussions in the physiology and function of the gastrointestinal tract. Epithelial sloughing and proliferation If immediate rejection of incoming larvae does not occur, expulsion of T. colubriformis from immune sheep is associated with transient loss of the jejunal epithelium at 4–5 days after intra-duodenal infection, effectively dislodging the larvae, which do not embed in the sub-epithelial mucosa (McClure et al., 1992). The epithelium is repaired within approximately 24 h. Fluid and electrolyte movement into lumen Mediators such as mast cell protease disrupt cell junctions, increasing mucosal permeability (Miller, 1996). This both increases the fluid content of the lumen and allows access of potentially protective host molecules to the worm. Thus the host mounts a wide range of immune and inflammatory responses to a challenge infection with worms, and probably no one response is essential or sufficient on its own for protection.

Influences of parasites on gastrointestinal function of host Effects of gastrointestinal parasitism on the digestive physiology of ruminants have been reviewed most recently by van Houtert and Sykes (1996) and Coop and Holmes (1996) with particular emphasis on quantitative aspects of digestion and absorption. Depression in food consumption is frequently, but not invariably, observed, and is

430

S.J. McClure et al.

dependent on the level of intake of infective larvae of small-intestinal (Steel et al., 1980) and abomasal (Symons et al., 1981) parasites. The mechanisms responsible are poorly understood although abdominal pain, gut inflammation, pH changes, changes in digesta flow rate, changes in protein to energy ratios of absorbed nutrients together with changes in cholecystokinin secretion have all been implicated (Symons, 1985), as well as neuropeptides and possibly some cytokines (van Houtert and Sykes, 1996). During continuous infections, appetite begins to return to normal after 10–12 weeks of larval intake. Although anorexia clearly has a profound effect on overall nutrient absorption, numerous pair-feeding studies have demonstrated that it does not entirely account for the impaired productivity of parasitized animals. Impaired uptake of nutrients at the site of small-intestinal parasite infections is attributable to deficiencies in digestive enzymes, proliferation of undifferentiated nonabsorbing cells on the villi, and reduced surface area due to villous atrophy (Steel and Symons, 1982). However, the overall capacity of the ruminant small intestine to digest and absorb protein appears to be generally unaffected by parasitism at proximal sites (Poppi et al., 1986; Kimambo et al., 1988a). The increased flow of nitrogen observed at the ileum during the first 12 weeks of T. colubriformis infection (Steel, 1974; Poppi et al., 1986; Kimambo et al., 1988a) and at the abomasum and terminal ileum during Teladorsagia circumcincta (Steel, 1978) and H. contortus (Rowe et al., 1988) infection have therefore been attributed to increased endogenous nitrogen secretion into the gastrointestinal tract. The principal sources of this endogenous nitrogen are increases at the site of infection in gastroenteric plasma and blood protein loss, turnover and loss into the gut of mucosal epithelia, and mucoprotein secretion by goblet cells. These changes appear to be associated in turn, not surprisingly, with an increased rate of protein synthesis by gastrointestinal tissues (Steel and Symons, 1982) such that there is a diversion of amino acids away from muscle, bone and wool growth into protein synthesis for processes essential for survival. Although often termed ‘pathophysiological’ these responses may indeed be an unavoidable component of the development of immune competence, and it is significant to note that they appear to peak at a time (around 12– 14 weeks) when manifestations of resistance mechanisms in weaner lambs, such as declining faecal egg count and the progressive expulsion of resident worms, are clearly evident (Steel et al., 1980; Poppi et al., 1986; Kimambo et al., 1988a).

Cost to the host of protective immunity The sequelae of physiological and immunological events responsible for the expulsion of resident adult worm populations in primary infections of young sheep correlates closely in time with maximal effects on liveweight gain and wool growth, particularly for small-intestinal infections (Steel et al., 1980; Kimambo et al., 1988a). Continued infection beyond this time is accompanied by a progressive recovery in productive performance and gastrointestinal function. Whether animals which have acquired a protective immunity to worm establishment experience a physiological and production cost during persistent exposure to infective larvae, because of hypersensitivity responses, has been of considerable interest since reduced wool growth and elevated plasma pepsinogen concentrations have been recorded in mature ‘resistant’ sheep under challenge (Barger 1982). Daily challenge of immune sheep for 10 weeks with T.

Host Resistance to Gastrointestinal Parasites

431

colubriformis larvae, following a 24-week parasite-free period, elicited a small increase in plasma nitrogen loss into the gut during the first 9 days, but no other change in nitrogen flow in digesta, digestibility or retention, and no excretion of worm eggs in faeces (Kimambo et al., 1988b). Other studies have recorded immediate but transient increases in enteric plasma loss 4–6 days after a single challenge of T. colubriformis given to lambs made solidly immune by vaccination with irradiated larvae (Steel et al., 1990), and in ‘immune’ ewes taken off contaminated pasture and given daily challenge with mixed trichostrongyle larvae, predominantly of Ostertagia circumcincta (Yakoob et al., 1983). However, in young sheep made highly resistant to infection by vaccination with irradiated larvae, continuous challenge with T. colubriformis larvae for 20 weeks had no effect on rates of liveweight gain or wool growth, despite evidence of immunological reactivity in terms of elevated serum antibody levels and total absence of faecal egg count (Wagland et al., 1982). Protective immunity in terms of larval exclusion and prevention of worm establishment therefore seems to be effected through changes in the composition of intestinal mucus, including the content of ‘larval migration inhibitory factor’, without any detectable metabolic or production cost (Kimambo and MacRae, 1988).

The non-specific enhancement of immunity Two features of immunity to worms, namely the role of inflammation in induction, and the expulsion of parasites via a pharmacologically-mediated non-specific effector response, raise the possibility of augmenting the levels of resistance to parasites. Feasible means to increase mucosal inflammation include concurrent (controlled) infections or infestations, immunomodulators or the judicious use of nutritional supplements. It is already appreciated that Trichostrongylus and particularly Ostertagia reduce the establishment of Haemonchus (Dobson and Barnes, 1995) and that genetically-selected resistant sheep mount more rapid and pronounced inflammatory responses to gastrointestinal nematodes (GIN) parasites (Windon, 1996). Irradiated Haemonchus or Trichostrongylus larval vaccines also induced about 30% protection against the heterologous parasite. Several other manipulations have been found to reduce parasitism: 1. Ground navy beans, given as 10% of pelleted diets and constituting 50% of digestible dietary protein, significantly (P < 0.05) reduced parasitism in sheep given T. colubriformis L3 or harbouring a mature infection (Ad) (Fig. 24.1). Sheep were fed the diet for up to 22 days from the day of infection (L3) or from 3 weeks into a primary infection (Ad), and in both situations, parasitism was reduced by > 60%. It is hypothesized that lectins within the navy beans increased mucosal inflammation in the same way that soybean diets can induce diarrhoea in feedlot cattle. 2. Variable levels of resistance, ranging from 35 to 50%, can be achieved against challenge with T. colubriformis by oral delivery of attenuated Salmonella typhimurium aro A (H.N. Brahmbatt, unpublished observations) or intraperitoneal inoculation of liposomes (D.L. Emery, unpublished observations). 3. Supplementaton of sheep diets with Mo also decreased levels of parasitism following challenge (Suttle et al., 1992a, b; McClure et al., 1999).

432

S.J. McClure et al. 30,000 Day 22

Day 15

Worm count

25,000

L3

Ad

L3

Ad

20,000 15,000 10,000 5,000 0

+



+



+



+



Beans (10% of diet; 50% of protein)

Fig. 24.1. A total of 32 sheep were fed a normal ration (⫺, 16 sheep) or a ration in which navy beans (10% wt/wt) provided 50% of dietary protein (+, 16 sheep). Eight sheep from each group were infected with 20,000 Trichostrongylus colubriformis L3 21 days prior to initiation of feeding, to examine effects on established adult worms (Ad), and the remaining sheep were infected with 20,000 T. colubriformis L3 coincidentally with the diet. Four sheep from each of the four treatment groups were killed for worm counts after 15 and 22 days of the two rations.

Although the ‘protective’ effects of these relatively diverse treatments could act directly or indirectly on the gut immune responses of the host, the exact mechanisms are not resolved at present.

Conclusions Unlike some internal parasites, GIN of ruminants have a simple life-cycle, and during their parasitic phase reside only in the gastrointestinal tract of the host. They therefore present a difficult immunological problem for the host, requiring protective immunity from an organ normally restrained from reacting to locally-presented molecules, and one with a good barrier between local immune system and pathogen. The gut immune responses associated with rejection of these parasites are physiologically complex and appear to contain much redundancy in that there is no evidence that any one mechanism is essential for protection. Protective immune responses against GIN parasites which are allergic in nature are also highly regulated within the local mucosal environment to avoid systemic sequelae. Mucosal immune responses come at considerable cost to the host in terms of cell and protein loss, reduced nutrient absorption and increased metabolic demand for nutrients; a cost exacerbated by the damage to gastrointestinal function due directly to the worm. It should be remembered that these mechanisms are subject to a number of influences. These include physiological factors such as age, liveweight and sex, and external factors such as nutrition and weaning stress. The nutritional requirements for optimal mucosal immunity appear to be greater than currently recognized, and additional to

Host Resistance to Gastrointestinal Parasites

433

those required to meet the demands of growth and production. The efficacy of the mucosal immune responses can also be enhanced by altering the local immunological environment, for example the degree of inflammation. Thus gut immunity is not an isolated component of ruminant physiology, but is highly integrated with the nervous, digestive and endocrine systems, and underlaid by a common cell biology regulating signalling, activation, metabolism, replication and differentiation. In such a situation, intervention which is directed towards addressing any single factor contributing to susceptibility will predictably have limited prospects of success.

References Adams, D.B. (1988) The effect of dexamethasone on a single and a superimposed infection with Haemonchus contortus in sheep. International Journal for Parasitology 18, 575–579. Adams, D.B. and Cripps, A.W. (1977) Cellular changes in the intestinal lymph of sheep infected with the enteric nematode, Trichostrongylus colubriformis. Australian Journal of Biology and Medical Science 55, 509–522. Barger, I.A. (1982) Helminth parasites and animal production. In: Symons, L.E.A., Donald, A.D. and Dineen, J.K. (eds) Biology and Control of Endoparasites. Academic Press, Sydney, pp. 133–155. Bendixsen, T., Emery, D.L. and Jones, W.O. (1995) The sensitization of mucosal mast cells during infections with Trichostrongylus colubriformis or Haemonchus contortus in sheep. International Journal for Parasitology 25, 741–748. Buddle, B.M., Jowett, G., Green, R.S., Douch, P.G.C. and Risdon, P.L. (1992) Association of blood eosinophilia with the expression of resistance in Romney lambs to nematodes. International Journal for Parasitology 22, 955–960. Coop, R.L. and Holmes, P.H. (1996) Nutrition and parasite interaction. International Journal for Parasitology 26, 951–962. Dawkins, H.J.S., Windon, R.G. and Eagleson, G.K. (1989) Eosinophil responses in sheep selected for high and low responsiveness to Trichostrongylus colubriformis. International Journal for Parasitology 19,199–205. Dineen, J.K., Gregg, P., Windon, R.G., Donald, A.D. and Kelly, J.D. (1977) The role of immunologically specific and non-specific components of resistance in cross-protection of intestinal nematodes. International Journal for Parasitology 7, 211–215. Dobson, R.J. and Barnes, E.H. (1995) Interaction between Ostertagia circumcincta and Haemonchus contortus infection in young lambs. International Journal for Parasitology 25, 495–501. Dobson, R.J., Waller, P.J. and Donald, A.D. (1990) Population dynamics of Trichostrongylus colubriformis in sheep: The effect of infection rate on the establishment of infective larvae and parasite fecundity. International Journal for Parasitology 20, 347–352. Douch, P.G.C., Harrison, G.B.L., Buchanan, L.L. and Greer, K.S. (1983) In vitro bioassay of sheep gastrointestinal mucus for nematode paralysing activity mediated by substances with some properties characteristic of SRS-A. International Journal for Parasitology 13, 207–212. Emery, D.L. and McClure, S.J. (1995) Studies on the inhibition of rejection of Trichostrongylus colubriformis larvae from immune sheep. International Journal for Parasitology 25, 761–764. Emery, D.L., Bendixsen, T. and McClure, S.J. (1991) The use of electroblotted antigens of Trichostrongylus colubriformis to induce proliferative responses in sensitized lymphocytes from sheep. International Journal for Parasitology 21, 179–185. Gill, H.S., Watson, D.L. and Brandon, M.R. (1993) Monoclonal antibody to CD4+ T cells abrogates genetic resistance to Hamonchus contortus in sheep. Immunology 78, 43–49.

434

S.J. McClure et al. Haig, D.M., Windon, R., Blackie, W., Brown, D. and Smith, W.D. (1989) Parasite-specific T cell responses of sheep following live infection with the gastric nematode Haemonchus contortus. Parasite Immunology 11, 463–477. Jackson, F., Miller, H.R.P., Newlands, G.J.F., Wright, S.E. and Hay, L.A. (1988) Immune exclusion of Haemonchus contortus larvae in sheep: dose dependency, steroid sensitivity and persistence of the response. Research in Veterinary Science 44, 320–323. Jones, W.O., Windon, R.G., Steel, J.W. and Outteridge, P.M. (1990) Histamine and leukotriene concentrations in duodenal tissue and mucus of lambs selected for high and low responsiveness to vaccination and challenge with Trichostrongylus colubriformis. International Journal for Parasitology 20, 1075–1079. Jones, W.O., Emery, D.L., McClure, S.J. and Wagland, B.M. (1994) Changes in inflammatory mediators and larval inhibitory activity in intestinal contents and mucus during primary and challenge infections of sheep with Trichostrongylus colubriformis. International Journal for Parasitology 24, 519–525. Karanu, F.N., McGuire, T., Davis, W., Besser, T.E. and Jasmer, D.P. (1997) CD4+ lymphocytes contribute to protective immunity induced in sheep and goats by Haemonchus contortus gut antigens. Parasite Immunology 19, 435–445. Kimambo, A.E. and MacRae, J.C. (1988) Measurement in vitro of a larval migration inhibitory factor in gastrointestinal mucus of sheep made resistant to the roundworm Trichostrongylus colubriformis. Veterinary Parasitology 28, 213–222. Kimambo, A.E., MacRae, J.C., Walker, A., Watt, C.F. and Coop, R.L. (1988a) Effect of prolonged infection with Trichostrongylus colubriformis on the performance and nitrogen metabolism of growing lambs. Veterinary Parasitology 28, 191–203. Kimambo, A.E., MacRae, J.C. and Dewey, P.J.S. (1988b) The effect of daily challenge with Trichostrongylus colubriformis larvae on the nutrition and performance of immunologically resistant sheep. Veterinary Parasitology 28, 205–212. McClure, S.J., Emery, D.L., Wagland, B.M. and Jones, W.O. (1992) A serial study of rejection of Trichostrongylus colubriformis by immune sheep. International Journal for Parasitology 22, 227–234. McClure, S.J., Davey, R.J., Lloyd, J.B. and Emery, D.L. (1995) Depletion of IFN-␥, CD8+ or Tcr ␥␦+ cells in vivo during primary infection with an enteric parasite (Trichostrongylus colubriformis) enhances protective immunity. Immunology and Cell Biology 73, 552–555. McClure, S.J., McClure, T.J. and Emery, D.L. (1999) Effects of molybdenum intake on primary infection and subsequent challenge by the nematode parasite Trichostrongylus colubriformis in Merino lambs. Research in Veterinary Science 67, 17–22. Miller, H.R.P. (1996) Mucosal mast cells and the allergic response against nematode parasites. Veterinary Immunology and Immunopathology 54, 331–336. Pfeffer, A., Douch, P.G.C., Shaw, R.J., Gatehouse, T.K., Rabel, B., Green, R.S., Shirer, C.L., Jonas, W.E. and Bisset, S. (1996) Sequential cellular and humoral responses in the abomasal mucosa and blood of Romney sheep dosed with Trichostrongylus axei. International Journal for Parasitology 26, 765–773. Poppi, D.P., MacRae, J.C., Brewer, A. and Coop, R.L. (1986) Nitrogen transactions in the digestive tract of lambs exposed to the intestinal parasite Trichostrongylus colubriformis. British Journal of Nutrition 55, 593–602. Rothwell, T.L.W. (1989) Immune expulsion of parasitic nematodes from the alimentary tract. International Journal for Parasitology 19, 139–168. Rowe, J.B., Nolan, J.V., Chaneet, G. de, Teleni, E. and Holmes, P.H. (1988) The effect of haemonchosis and blood loss into the abomasum on digestion in sheep. British Journal of Nutrition 59, 125–139. Shaw, R.J., Gatehouse, T.K. and McNeill, M.M. (1998) Serum IgE responses during primary and challenge infections of sheep with Trichostrongylus colubriformis. International Journal for Parasitology 28, 293–302.

Host Resistance to Gastrointestinal Parasites

435

Smith, W.D. (1993) Protection in lambs immunized with Haemonchus contortus gut membrane proteins. Research in Veterinary Science 54, 94–101. Smith, W.D., Jackson, F., Jackson, E., Williams, J. and Miller, H.R.P. (1984) Manifestations of resistance to ovine ostertagiasis associated with immunological responses in the gastric lymph. Journal of Comparative Pathology 94, 591–602. Smith, W.D., Jackson, F., Jackson, E. and Williams, J. (1985) Age immunity to Ostertagia circumcincta: comparison of the local immune responses of 4 and 10-month-old lambs. Journal of Comparative Pathology 95, 235–245. Stear, M.J., Park, M. and Bishop, S.C. (1996) The key components of resistance to Ostertagia circumcincta in lambs. Parasitology Today 12, 438–441. Steel, J.W. (1974) Pathophysiology of gastrointestinal nematode infections in the ruminant. Proceedings of the Australian Society of Animal Production 10, 139–147. Steel, J.W. (1978) Inter-relationships between gastrointestinal helminth infection, nutrition and impaired productivity in the ruminant. In: Farrell, D.J. (ed.) Recent Advances in Animal Nutrition in Australia. The University of New England, Armidale, Australia, pp. 98–109. Steel, J.W. and Symons, L.E.A. (1982) Nitrogen metabolism in nematodosis of sheep in relation to productivity. In: Symons, L.E.A., Donald, A.D. and Dineen, J.K. (eds) Biology and Control of Endoparasites. Academic Press, Sydney, pp. 235–256. Steel, J.W., Symons, L.E.A. and Jones, W.O. (1980) Effects of level of larval intake on the productivity and physiological and metabolic responses of lambs infected with Trichostrongylus colubriformis. Australian Journal of Agricultural Research 31, 821–838. Steel, J.W., Jones, W.O. and Wagland, B.M. (1990) The response of immune sheep to challenge with Trichostrongylus colubriformis: enteric plasma loss and secretion of biogenic amines. International Journal for Parasitology 20, 1067–1073. Stewart, M.J., Emery, D.L. and McClure, S.J. (1995a) Effects of four neuropeptides in lymphocytes from sheep. Regional Immunology 6, 264–269. Stewart, M.J., McClure, S.J., Rothwell, T.L.W. and Emery, D.L. (1995b) Changes in the enteric innervation of the sheep during primary or secondary immune responses to the nematode Trichostrongylus colubriformis. Regional Immunology 6, 460–470. Stewart, M.J., Emery, D.L., McClure, S.J. and Bendixsen, T. (1996) The effects of four neuropeptides on the degranulation of mucosal mast cells from sheep. Immunology and Cell Biology 74, 255–257. Symons, L.E.A. (1985) Anorexia; occurrence, pathophysiology and possible causes in parasitic infections. In: Baker, J.R. and Muller, R. (eds) Advances in Parasitology, Vol. 24. Academic Press, London, pp. 103–133. Symons, L.E.A., Steel, J.W. and Jones, W.O. (1981) Effects of level of larval intake on the productivity and physiological and metabolic responses of lambs infected with Ostertagia circumcincta. Australian Journal of Agricultural Research 32, 139–148. Suttle, N.F., Knox, K.W., Angus, K.W., Jackson, F. and Coop, R.L. (1992a) Effects of dietary molybdenum on nematode and host during Haemonchus contortus infection in lambs. Research in Veterinary Science 52, 230–235. Suttle, N.F., Knox, K.W., Jackson, F., Coop, R.L. and Angus, K.W. (1992b) Effects of dietary molybdenum on nematode and host during Trichostrongylus vitrinus infection in lambs. Research in Veterinary Science 52, 224–229. Tremain, S.A. and Emery, D.L. (1994) Myoelectric activity of the small intestine during parasite rejection. 30th Annual Scientific Meeting of the Australian Society for Parisitology. Australian Society for Parasitology, Sydney, p. 43. Vallance, B.A. and Collins, S.M. (1998) The effect of nematode infection upon intestinal smooth muscle function. Parasite Immunology 20, 249–253. van Houtert, M.F.J. and Sykes, A.R. (1996) Implications of nutrition for the ability of ruminants to withstand gastrointestinal nematode infections. International Journal for Parasitology 11, 1151–1168.

436

S.J. McClure et al. Wagland, B.M., Steel, J.W. and Dineen, J.K. (1982) Effect of continuous challenge with Trichostrongylus colubriformis on liveweight gain and wool growth of resistant sheep. Proceedings of the Australian Society of Animal Production 14, 642. Wagland, B.M., Emery, D.L. and McClure, S.J. (1996) Studies on the host–parasite relationship between Trichostrongylus colubriformis and susceptible and resistant sheep. International Journal for Parasitology 26, 1279–1286. Windon, R.G. (1996) Genetic control of resistance to helminths in sheep. Veterinary Immunology and Immunopathology 54, 245–254. Yakoob, A., Holmes, P.H. and Armour, J. (1983) Pathophysiology of gastrointestinal trichostrongyles in sheep: plasma losses and changes in plasma pepsinogen levels associated with parasite challenge of immune animals. Research in Veterinary Science 34, 305–309.

25

Host Resistance to Fleece Rot and Blowfly Strike

I.G. COLDITZ1 AND R.L. TELLAM2 1CSIRO

Animal Production, Pastoral Research Laboratory, Armidale, New South Wales, Australia; 2CSIRO Tropical Agriculture, Longpocket Laboratory, Indooroopilly, Queensland, Australia

Introduction Interest in resistance of sheep and cattle to external parasites arose soon after emergence of blowfly strike and cattle tick as diseases of major economic importance in Australia. It is most likely that blowflies were introduced to Australia on several occasions: Lucilia sericata from Britain during the early days of European settlement; and Lucilia cuprina from South Africa or India during the latter stages of the 19th century (Gilruth et al., 1933). Blowfly strike was sporadic until 1903 when it became widespread in New South Wales and Victoria, extending throughout the range of sheep over the next few years (Gilruth et al., 1933). Emergence of the disease coincided with introductions of American Merinos from Vermont, which had a high grease content in the fleece and pronounced skin wrinkle (Belschner, 1966). With the realization that body conformation was a predisposing factor for susceptibility to blowfly strike (Seddon et al., 1931a) and that these predisposing traits were heritable (Seddon et al., 1931b), interest turned to breeding sheep resistant to the disease and to surgical removal of skin folds in the breech (mulesing). Wetting of fleece was found to induce bacterial growth (Stuart, 1894), which in turn induced green or red discoloration of wool (Seddon and McGrath, 1929) and dermatitis (Bull, 1931). These studies led to the conclusion that predisposing conditions for blowfly strike are wool characteristics and body conformations that favour, principally through susceptibility to wetting and resultant dermatitis, attraction of flies, oviposition and subsequent nutrition of larvae (Gilruth et al., 1933). Thus fleece rot and blowfly strike occur as a disease complex, with the strongest interdependence when blowfly strike occurs over the shoulders, back and flanks (body strike). This early work provided the basis for the next 70 years of research into resistance to blowfly strike. Major areas of research have been genetic studies on direct and indirect selection for resistance to blowfly strike, mechanisms of resistance in selected flocks, innate immunity, acquired immunity and vaccines. Progress in these areas will be considered in turn. © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

437

438

I.G. Colditz and R.L. Tellam

Direct and indirect selection for resistance to blowfly strike Hayman (1953) confirmed Seddon’s observations on heritability of resistance to fleece rot. Quantitatively, McGuirk and Atkins (1984) estimated a heritability of susceptibility to fleece rot of 0.40 in unselected medium-wool Peppin Merinos from 15 years of field data at Trangie Agricultural Research Centre in New South Wales. Heritability of susceptibility to blowfly strike in the same environment was between 0.26 and 0.37 (Gilmour and Raadsma, 1986; Raadsma, 1991). Experimental conditions for inducing fleece rot and fly strike were developed at Trangie using artificial wetting by overhead sprinklers of sheep temporarily housed indoors (McGuirk et al., 1978). Fleece rot and subsequent body strike resulting from artificial wetting provided the basis for establishing two selection lines with resistance and susceptibility to the disease complex (McGuirk et al., 1978). These selection lines have been the subjects of intensive study of both wool characteristics and physiological responses associated with resistance. Heritability of susceptibility to fleece rot under artificial wetting was comparable (Raadsma et al., 1989) to that seen under field conditions (Raadsma and Rogan, 1987). Following 17 years of selection, the lines had diverged at an annual rate of 2.8% for natural fleece rot and 0.4% for natural body strike (Mortimer et al., 1998). With the absence of a random-bred control line in this study it is unclear whether there has been a greater impact of selection on resistance or susceptibility to the disease complex. The sporadic occurrence in many of the grazing zones of Australia of weather conditions that induce fleece rot and that accompany fly waves stimulated research on correlated wool traits that could provide the basis for indirect selection for resistance. Wettability of the fleece (Raadsma, 1989), objectively measured wool colour (Raadsma and Wilkinson, 1990) and fibre characteristics (Raadsma, 1993) are genetically correlated with resistance to the disease complex. Despite these findings there has been limited selection for these correlated traits in Merino breeding programmes other than through the long-held husbandry practice of culling susceptible sheep on the basis of conformational faults, subjectively assessed wool characteristics and wool discoloration associated with fleece rot.

Mechanisms of resistance in the Trangie selection flock Freshly hatched larvae feed on proteins in exudates associated with bacterial dermatitis (Gilruth et al., 1933; Sandeman et al., 1987). Skin inflammatory responses were therefore examined to determine whether underlying differences in plasma leakage could account for differences between the selection lines in susceptibility to blowfly strike. Leakage of radiolabelled albumin from the vascular compartment into skin was measured following intradermal injection of histamine, bradykinin, activated complement, platelet-activating factor or serotonin. Plasma leakage induced by activated complement was greater in susceptible sheep, and for all mediators there was a tendency for plasma leakage to be greater in susceptible sheep (Colditz et al., 1992a). This finding suggested that a smaller plasma leakage response following release of endogenous permeability mediators might contribute to resistance to blowfly strike and is in accord with the nutritive role of blood proteins for larvae. Serum complement C3 levels, however, did not differ between lines (O’Meara and Raadsma, 1995), suggesting that differ-

Host Resistance to Fleece Rot and Blowfly Strike

439

ences might lie at the level of receptor number or function controlling vascular permeability rather than in supply of mediator. Contrary to this finding is the observation that when excretory and secretory products from larvae were injected intradermally, there was a greater wheal response in resistant animals (O’Meara et al., 1992). The wheal response to excretory and secretory products has also been correlated with resistance to fleece rot and blowfly strike in unselected sheep (Broadmeadow, 1988) and has been proposed as a selection marker for resistance to the disease complex (Raadsma et al., 1992). Cellular components of the inflammatory response in skin and resident leucocytes in skin have also been examined in the resistant and susceptible lines. Intense neutrophil accumulation is a feature of the dermatitis that accompanies fleece rot and fly strike (Bull, 1931; Burrell et al., 1982; Bowles et al., 1992) and there is extensive production of inflammatory cytokines interleukin (IL)-1␣, IL-1␤, IL-6 and IL-8 (Elhay et al., 1994) at the wound site. Neutrophil accumulation in response to intradermal injection of ovine IL-1␤, human IL-8, human IL-1␣, human tumour necrosis factor (TNF)-␣ activated complement, leukotriene B4 and endotoxin from Pseudomonas aeruginosa did not differ between lines (Colditz et al., 1994). Leucocyte populations in untreated skin and in skin sites collected 6 h following intradermal injection of TNF-␣ were examined by immunohistology. WC1+ lymphocytes and eosinophils were more prevalent in skin of susceptible sheep and IgE+ cells (probably mast cells) were more prevalent in skin of resistant animals (Colditz et al., 1994). In an independent study (Nesa, 1994) mast cells were also found to be more prevalent in skin from animals of the resistant genotype. This difference in the prevalence of mast cells may underlie the greater wheal response to intradermal injection of larval antigens in resistant sheep noted above. Furthermore, in an examination of the IgE gene, Engwerda et al. (1998) found a restriction fragment length polymorphism between resistant and susceptible lines. Together these findings suggest that the role of mast cells in resistance to blowfly strike deserves further attention. The response of resistant and susceptible sheep to artificial infestation with larvae has been examined in two experiments. O’Meara et al. (1995) found greater exudation of serum proteins onto the skin surface during the first 12 h of infestation in resistant animals, which is in accord with greater skin wheal responses seen in these animals. In the second experiment, the effects of lymphocyte and interferon (IFN)-␥ depletion on larval establishment and growth have been studied by treatment of sheep with monoclonal antibodies. In contrast to results seen with the internal parasite, Trichostrongylus colubriformis, where depletion of CD8+, WC1+ lymphocytes or IFN-␥ enhanced resistance to infection (McClure et al., 1996), depletion of CD4+, CD8+, WC1+ lymphocytes or IFN-␥ had no effect on establishment or growth of blowfly larvae (Colditz et al., 1996). Importantly in both artificial infestations, there were no differences between the resistant and susceptible lines in the establishment or growth of larvae. Lymphocyte subsets in blood do not differ between resistant and susceptible lines (Colditz et al., 1996; McColl et al., 1997). Following intravenous challenge with endotoxin from P. aeruginosa there are higher neutrophil counts and monocyte counts in the resistant line (I. Colditz, unpublished findings). Antibody responses to intradermal injection of P. aeruginosa antigens are generally greater in resistant sheep (Chin and Watts, 1991; Gogolewski et al., 1996); however, no significant differences between the lines were found in the antibody response to L. cuprina antigens (O’Meara et al.,

I.G. Colditz and R.L. Tellam

440

1997). Sheep from the resistant line develop higher titres of antibody to P. aeruginosa antigens during simulated fleece rot conditions when live cultures of P. aeruginosa are applied epicutaneously to wetted skin (Chin and Watts, 1991). This may result from differences in skin characteristics between the lines resulting in greater uptake of antigens through skin in the resistant animals, or alternatively to differences between lines in immunological recognition and response to P. aeruginosa antigens. Systemically administered P. aeruginosa vaccines can confer resistance to fleece rot (Burrell, 1985) so the differences in antibody titres noted by Chin and Watts (1991) and Gogolewski et al. (1996) may well contribute to the differences in prevalence and severity of fleece rot seen between the selection lines. Taken together, these extensive studies on the Trangie resistant and susceptible lines suggest that reactivity of skin may contribute to resistance but may play a subsidiary role to wool characteristics that predispose to fleece rot.

Innate immunity in unselected sheep In a review of fleece rot and body strike, McGuirk and Watts (1983) concluded that there appear to be three barriers providing resistance to blowfly strike: wool, skin and the immune system. Wool characteristics have been mentioned briefly above in relation to the Trangie selection flocks. Physical skin characteristics affecting resistance such as composition of fats and wettability have received less attention than inflammatory reactions in skin. The host response to fleece rot and fly strike results in production of high levels of inflammatory cytokines in skin (Elhay et al., 1994). We therefore examined the direct toxic effects of these endogenous mediators on larvae in an in vitro feeding assay (Colditz and Eisemann, 1994). At concentrations that exhibit inflammatory activity in skin (Colditz, 1991; Colditz and Watson, 1992; Mulder and Colditz, 1993), histamine, bradykinin, platelet-activating factor, serotonin, activated complement, leukotriene B4, IL-8, IL-1␤, TNF-␣ and IFN-␥ all failed to affect larval growth. Antibody to peritrophic membrane antigens was an exception, inducing dose-dependent inhibition of larval growth. Protease inhibitors in plasma inhibit growth of larvae in vitro (Bowles et al., 1990) and anti-inflammatory drugs retard larval growth (O’Sullivan et al., 1984). These findings suggest that, with the probable exception of antibody, the promotion of larval growth by nutrients is greater than the concomitant inhibition induced by protease inhibitors present in the exudate (O’Meara et al., 1995). In accord with the finding in the Trangie selection lines noted above, IgE+ cells are more prevalent in skin of sheep with no history of fleece rot than in sheep with a record of fleece rot (Colditz et al., 1994). We recently examined factors contributing to the break in wool fibres that accompanies strike. Blowfly strike was induced in a group of five, 18-month-old Merino wethers by application of approximately 500 freshly hatched larvae per day for 8 days to an abraded skin site on the flank of each sheep. Daily feed intakes were recorded and pair-fed sheep were offered the quantity of feed consumed by their pair on the preceding day. A third group of uninfested controls was fed ad libitum for the duration of the experiment. Severe strike was induced in four of the infested sheep with one sheep exhibiting only mild strike. Average daily gain was significantly lower in infested sheep than control sheep during week one of infestation. Rectal temperature was elevated in infested sheep from the first day of infestation until 3 days after the last application of

Host Resistance to Fleece Rot and Blowfly Strike

441

infective larvae, which coincided with the treatment of all sheep with the organophosphate Diazinon to terminate infestations. Plasma cortisol levels were elevated in infested sheep from day 2 to day 6 of infestation. The cytokines IL-1␤, IL-6, IL-8 and TNF-␣ were assayed in plasma. IL-6 was elevated from day 2 to day 6 of infestation, whereas no significant differences were observed between treatment groups for concentrations of the other cytokines. Longitudinal growth of wool fibres did not differ between groups, though there was a trend towards less growth in struck and pair-fed sheep than in controls. Staple strength was significantly lower in struck sheep than in control and pair-fed sheep, but did not differ between sites near to and distant from the strike lesion. Taken together these findings suggest that the host response to infestation has systemic consequences that lead to reduced fibre strength throughout the fleece, and that the reduced feed intake accompanying blowfly strike is not primarily responsible for reduced fibre strength. Cortisol and IL-6 are implicated as systemic mediators that contribute to reduced fibre strength; however, the interdependence, pleiotropism and redundancy of mediators associated with stress and inflammatory responses make it unlikely that a single mediator will be identified as accountable for the effect of blowfly strike on fibre strength.

Acquired immunity in unselected sheep There have been several studies on the acquisition of immunity following repeated experimental infestation of sheep. Sheep rapidly produce antibodies to larval antigens (O’Donnell et al., 1980; Sandeman et al., 1985; Eisemann et al., 1990; Seaton et al., 1992) and there are dramatic changes in leucocyte populations at the site of infestation (Bowles et al., 1992) and in draining lymph (Bowles et al., 1994). There is, however, very limited effect of such sensitization from experimental infestations on survival and growth of larvae during subsequent challenge. It is noteworthy that larvae exert an immunosuppressive effect on the host, which has been attributed, in part, to excretion of ammonia (Guerrini, 1998). While lymphocyte depletion in the Trangie selection lines failed to affect larval establishment and growth, the effect of immunosuppressive treatments on growth of larvae in unselected sheep has not been examined. An infiltrate of CD4+, WC1+ and CD1+ leucocytes is seen at sites of infestation following immunization with larval antigens (Bowles et al., 1996). However, leucocyte depletion studies to determine the contribution of this infiltrate to reduced larval survival in immunized sheep have not been performed. Thus the role of cellular responses in resistance to blowfly strike remains an open question.

Development of blowfly vaccines The failure of repeated larval infestations to promote strong immunity and the success of the hidden antigen approach in the development of the cattle tick vaccine (Willadsen et al., 1989; Tellam et al., 1992) led to studies on gut antigens as candidates for a blowfly vaccine. It was argued that it may be possible to induce a humoral immune response in sheep that produced specific antibodies that, after ingestion by the larvae, bound to the larval gut. These antibodies might have the potential of disrupting

442

I.G. Colditz and R.L. Tellam

the normal functions of the gut in a manner deleterious to the larvae. The strategy relies on the ingestion of sufficient quantities of biologically active antibodies. It was found that larvae feeding on sheep ingest substantial quantities of functional antibodies (Eisemann et al., 1993). High concentrations of ingested antibody persist throughout the larval foregut and the anterior midgut. The quantity of antibody decreases sharply posterior to this region as a consequence of proteolysis. In contrast, only very small quantities of ingested antibody penetrate the wall of the midgut to reach the haemolymph and thence internal tissues. The most promising ‘concealed’ targets for immunological attack therefore would appear to lie in the anterior midgut. This region is exposed to ingested antibodies and is not protected by an impermeable layer of cuticle as is present in the foregut and hindgut. However, the midgut region is protected by a semi-permeable matrix, the peritrophic membrane, which is thought to have a central role in the facilitation of the digestive process in the gut and protection of the insect from invasion by bacteria. East et al. (1993) used this knowledge in testing crude extracts of peritrophic membrane in sheep vaccination trials. Sera from vaccinated sheep strongly inhibited larval growth in an in vitro feeding bioassay (East et al., 1993). Moreover, significant although weaker effects on larval growth were observed using bioassays directly on the backs of sheep. Vaccination trials in sheep were then used as an assay to guide the purification of specific peritrophic membrane antigens. This process, which involved many steps, led to the identification of several potential vaccine antigens but particularly the glycoprotein, peritrophin-95 (Casu et al., 1997). Vaccination of sheep with this antigen led to greater than 50% reduction in larval weights as measured by an in vitro larval feeding bioassay. The cDNA coding for this protein has been sequenced and various recombinant proteins produced. These were tested in vaccination trials which showed significant, albeit weak, anti-larval effects. Further studies of both the structure of the recombinant antigens and the nature of the antibody response to the native protein demonstrated that the nature of the oligosaccharides attached to this glycoprotein is an important determinant of the efficacy of the antilarval immune response induced by this antigen. Further studies are under way to ensure the appropriateness of the oligosaccharides attached to the recombinant proteins. Isolation of specific anti-peritrophin-95 antibodies and their re-constitution into control sera transferred the antilarval growth activity when measured in in vitro feeding bioassays. Moreover, higher concentrations of antibody resulted in more severe effects on larval growth. There was also a direct correlation between antibody titre from a number of sheep and the degree of larval growth inhibition measured from each serum in an in vitro feeding bioassay (R.L. Tellam and C.H. Eisemann, unpublished results). These experiments and others proved that the antilarval effect was mediated by antibody and that complement was not required. Examination of the peritrophic membrane from larvae feeding on sera from sheep vaccinated with peritrophin-95 showed that the luminal side of the peritrophic membrane was lined with a new layer of material of undefined composition. This layer was impermeable to 6 nm colloidal gold particles that normally freely diffused across the peritrophic membrane. Presumably, the layer inhibited the movement of nutrients into the digestive epithelia and resulted in the starvation of the larvae (R.L. Tellam and C.H. Eisemann, unpublished results). A vaccine based on the peritrophin-95 antigen is being developed but will require greater efficacy to be effective in the field. Combinations of this antigen with antigens possibly

Host Resistance to Fleece Rot and Blowfly Strike

443

involved in larval establishment on the host epidermis (Bowles et al., 1996) may result in a more efficacious vaccine.

Vaccine delivery With the realization that antibody held promise as a defence mechanism for protection against blowfly larvae, we turned our attention to investigating immunization protocols for elevating antibody concentrations in skin. Colleagues in our laboratory developed an apparatus for inducing transudation of interstitial fluid by applying a vacuum of ⫺25 kPa for 90 min to the skin surface (Watson et al., 1992). Up to 40 ␮l of interstitial fluid collects on the skin surface during this procedure (Colditz et al., 1992b). The ratio of 125I-IgG1 to 111In-transferrin in blood was compared with the ratio in normal skin and at skin sites receiving the vascular permeability mediators histamine, bradykinin, activated complement, platelet-activating factor or serotonin to determine whether there is preferential transfer of IgG1 to these extravascular sites. No evidence was obtained for selective transport of IgG1 into the dermis or onto the skin surface (Colditz et al., 1992b). Immunoglobulins can be detected on the surface of sheep skin in low concentrations (Lloyd et al., 1979) and our results suggest that for IgG1 this occurs by filtration rather than by selective transport. It is noteworthy, however, that IgA is selectively transported across sebaceous glands and sweat glands in man (Gebhart and Metze, 1990) and can be detected in skin washings from sheep following infection with Dermatophilus congolensis (Sutherland et al., 1987). The potential to induce local antibody responses in sheep by topical application of antigens was examined in a number of experiments (Colditz and Watson, 1993). Antigen delivery via a Panjet vaccination gun was more effective than topical application of antigen onto skin. Adjuvation of the experimental antigen ovalbumin with immunostimulating complexes (ISCOMs) resulted in further elevation of the antibody response. Antibody titres were higher at locally immunized skin sites than at nonimmunized skin sites of the same animals. Isotype analysis of antibody present in vacuum transudates indicated that immunoglobulin (Ig)M and IgG1 were elevated at the locally immunized skin sites, IgG2 antibody did not differ between sites and IgA antibody was not detected. Numbers of CD4+, CD8+ and WC1+ lymphocytes were elevated in the dermis at sites of local immunization (I.G. Colditz, D.L. Watson and S.J. McClure 1999, unpublished observations). Similar results have been obtained with recombinant peritrophic membrane antigens PM44, PM48 and PM95 from L. cuprina (I.G. Colditz, C.H. Eisemann and R.L. Tellam, unpublished observations). Together these findings demonstrate the potential for elevating the concentration of antibody in skin by local application of antigen. It has recently been shown that antigens combined with cholera toxin can induce high titres of systemic antibody when applied to unbroken skin in mice (Glenn et al., 1998a,b). No local gross or histological reactions accompanied this transcutaneous immunization, which protected mice from a lethal intranasal challenge with cholera toxin. We have recently observed a systemic antibody response in sheep following topical application of 100 ␮g of cholera toxin to skin (R.B. Cope and I.G. Colditz, unpublished observations). Interestingly, during fleece rot and dermatophilosis, sheep develop systemic antibody responses to these non-invasive infectious organisms, although

I.G. Colditz and R.L. Tellam

444

sensitization in this natural disease setting is generally thought to be not protective (Sutherland et al., 1987). The adjuvation effect of cholera toxin applied topically to skin may provide a new strategy for elevating antibody responses to vaccines against dermal pathogens and external parasites of sheep and cattle. Furthermore, the potential of cholera toxin to induce local antibody responses in skin also deserves investigation. Percutaneous delivery of vaccines should be very attractive to livestock industries due to ease of application, lower requirement for animal restraint, avoidance of carcass and hide damage and possibly no need for asepsis.

Conclusions A century’s research on blowfly strike has yielded a great deal of information on the pathogenesis of infestation and the nature of the host–parasite interaction. Despite these advances, the two major goals for research in this field, development (or selection) of resistant animals and the induction of resistance by vaccination, remain unattained. In view of the sporadic occurrence of conditions that permit direct selection for resistance, a major shortcoming of research to date is the absence of a reliable trait for indirect selection for resistance. As discussed by K.A. Ward in Chapter 21, transgenesis may yet yield novel solutions to the quest for resistant animals. For vaccines, solutions to the problem of invoking protective immunity may come not from harnessing components of naturally acquired resistance but by broaching the adapted host parasite interface by vaccination with evolutionarily naïve, concealed antigens.

Acknowledgements The collaboration of S. Walkden-Brown, B. Crook, B. Daley and C. Eisemann in studies of the effect of blowfly strike on fibre strength is gratefully acknowledged. This work was supported in part by Australian woolgrowers through the Australian Wool Research and Promotion Organisation and the L.W. Bett trust.

References Belschner, H.G. (1966) Sheep Management and Diseases, 8th edn. Angus and Robertson, Sydney, 814 pp. Bowles, V.M., Feehan, J.P. and Sandeman, R.M. (1990) Sheep plasma protease inhibitors influencing protease activity and growth of Lucilia cuprina larvae in vitro. International Journal for Parasitology 20, 169–174. Bowles, V.M., Grey, S.T. and Brandon, M.R. (1992) Cellular immune responses in skin of sheep infected with larvae of Lucilia cuprina, the sheep blowfly. Veterinary Parasitology 44, 151–162. Bowles, V.M., Meeusen, E.N.T., Chandler, K., Verhagen, A., Nash, A.D. and Brandon, M.R. (1994) The immune response of sheep infected with larvae of the sheep blowfly Lucilia cuprina monitored via efferent lymph. Veterinary Immunology and Immunopathology 40, 341–352. Bowles, V.M., Meeusen, E.N.T., Young, A.R., Andews, A.E., Nash, A.D. and Brandon, M.R.

Host Resistance to Fleece Rot and Blowfly Strike

445

(1996) Vaccination of sheep against larvae of the sheep blowfly (Lucilia cuprina). Vaccine 14, 1347–1352. Broadmeadow, M.E. (1988) Studies on resistance of sheep to blowfly strike. Wool Research and Development Council Project DAQ 21P Final Report, Queensland Department of Primary Industries, Brisbane, pp. 25. Bull, L.B. (1931) Some observations on dermatitis of the folds in the breech of sheep and its possible relationship to blowfly strike. Australian Veterinary Journal 7, 143–148. Burrell, D.H. (1985) Immunisation of sheep against experimental Pseudomonas aeruginosa dermatitis and fleece rot associated body strike. Australian Veterinary Journal 62, 55–57. Burrell, D.H., Merritt, G.C., Watts, J.E. and Walker, K.H. (1982) Experimental production of dermatitis in sheep with Pseudomonas aeruginosa. Australian Veterinary Journal 59, 140–144. Casu, R., Eisemann, C., Pearson, R., Riding, G., East, I., Donaldson, A., Cadogan, L. and Tellam, R. (1997) Antibody-mediated inhibition of the growth of larvae from an insect causing cutaneous myiasis in a mammalian host. Proceedings of the National Academy of Sciences USA 94, 8939–8944. Chin, J.E. and Watts, J.E. (1991) Dermal and serological response against Pseudomonas aeruginosa in sheep bred for resistance and susceptibility to fleece-rot. Australian Veterinary Journal 68, 28–31. Colditz, I.G. (1991) The induction of plasma leakage in skin by histamine, bradykinin, activated complement, platelet-activating factor and serotonin. Immunology and Cell Biology 69, 215–219. Colditz, I.G. and Eisemann, C.H. (1994) The effect of immune and inflammatory mediators on growth of Lucilia cuprina larvae in vitro. International Journal for Parasitology 24, 401–403. Colditz, I.G. and Watson, D.L. (1992) The effect of cytokines and chemotactic agonists on the migration of T lymphocytes into skin. Immunology 76, 272–278. Colditz, I.G. and Watson, D.L. (1993) Immunisation to induce antibody production in skin of sheep. Proceedings of Vaccines 93, Lorne House, Victoria (abstract J94). Colditz, I.G., Woolaston, R.R., Lax, J. and Mortimer, S.I. (1992a) Plasma leakage in skin of sheep selected for resistance or susceptibility to fleece rot and fly strike. Parasite Immunology 14, 587–594. Colditz, I.G., Altmann, K.G. and Watson, D.L. (1992b) Intradermal and percutaneous transudation of IgG1 and transferrin in sheep. Immunology and Cell Biology 70, 323–327. Colditz, I.G., Lax, J., Mortimer, S.I., Clarke, R.A. and Beh, K.J. (1994) Cellular inflammatory responses in skin of sheep selected for resistance or susceptibility to fleece rot and fly strike. Parasite Immunology 16, 289–296. Colditz, I.G., Eisemann, C.H., Tellam, R.L., McClure, S.J., Mortimer, S.I. and Husband, A.J. (1996) Growth of Lucilia cuprina larvae following treatment of sheep divergently selected for fleece rot and fly strike with monoclonal antibodies to T lymphocyte subsets and interferon gamma. International Journal for Parasitology 26, 775–782. East, I.J., Fitzgerald, C.G., Pearson, R.D., Donaldson, R.A., Vuocolo, T., Cadogan, L.C., Tellam, R.L. and Eisemann, C.H. (1993) Lucilia cuprina: inhibition of larval growth induced by immunization of host sheep with extracts of larval peritrophic membrane. International Journal for Parasitology 23, 221. Eisemann, C.H., Johnston, L.A.Y., Broadmeadow, M., O’Sullivan, B.M., Donaldson, R.A., Pearson, R.D., Vuocolo, T. and Kerr, J.D. (1990) Acquired resistance of sheep to larvae of Lucilia cuprina, assessed in vivo and in vitro. International Journal for Parasitology 20, 299–305. Eisemann, C.H., Pearson, R.D., Donaldson, R.A., Cadogan, L.C. and Vuocolo, T. (1993) Uptake and fate of specific antibody in feeding larvae of the sheep blowfly, Lucilia cuprina. Medical and Veterinary Entomology 7, 177–185. Elhay, M.J., Hanrahan, C.F., Bowles, V.M., Seow, H., Andrews, A.E. and Nash, A.D. (1994)

446

I.G. Colditz and R.L. Tellam Cytokine mRNA expression in skin in response to ectoparasite infection. Parasite Immunology 16, 451–461. Engwerda, C.R., Dale, C.J. and Sandeman, R.M. (1998) IgE, TNF␣, IL1␤, IL4 and IFNg gene polymorphisms in sheep selected for resistance to fleece rot and flystrike. International Journal for Parasitology 26, 787–791. Gebhart, W. and Metze, D. (1990) Secretory immunoglobulins. In: Bos, J.D. (ed.) Skin Immune System. CRC Press, Boca Raton, pp. 197–206. Gilmour, A.R. and Raadsma, H.W. (1986) Estimating genetic variation and covariation for flystrike incidence in Australian Merino sheep on an underlying normal scale. 3rd World Congress on Genetics applied to Livestock Production, Lincoln, Nebraska, USA, July 16–22, 1986. XII. Biotechnology, selection experiments, parameter estimation, design of breeding systems, management of genetic resources. p. 460. Gilruth, J.A., Tillyard, R.J., Seddon, H.R., Gurney, W.B. and Mackerras, I.M. (1933) The sheep blowfly problem, Report No. 1 of the Joint Blowfly Committee. NSW Department of Agriculture Science Bulletin 40, 1–136. Glenn, G.M., Rao, M., Matyas, G.R. and Alving, C.R. (1998a) Skin immunization made possible by cholera toxin. Nature 391, 851. Glenn, G.M., Scharton-Kersten, T., Vassell, R., Mallett, C.P., Hale, T.L. and Alving, C.R. (1998b) Cutting Edge: Transcutaneous immunisation with cholera toxin protects mice against lethal mucosal toxin challenge. Journal of Immunology 161, 3211–3214. Gogolewski, R.P., Nicholls, P.J., Mortimer, S.I., Mackintosh, J.A., Nesa, M., Ly, W. and Chin, J.C. (1996) Serological responses against Pseudomonas aeruginosa in Merino sheep bred for resistance or susceptibility to fleece rot and body strike. Australian Journal of Agricultural Research 47, 917–926. Guerrini, V.H. (1998) Excretion of ammonia by Lucilia cuprina larvae suppresses immunity in sheep. Veterinary Immunology and Immunopathology 56, 311–317. Hayman, R.H. (1953) Studies on fleece-rot in sheep. Australian Journal of Agricultural Research 4, 430–468. Lloyd, D.H., McEwan Jenkinson, D. and Mabon, R.M. (1979) Location of immunoglobulins in skin of cattle and sheep. Research in Veterinary Science 26, 47–52. McClure, S.J., Davey, R.L., Emery, D.L., Colditz, I.G. and Lloyd, J.B. (1996) In vivo depletion of T-cells and cytokines during primary exposure of sheep to parasites. Veterinary Immunology and Immunopathology 54, 83–90. McColl, K.A., Gogolewski, R.P. and Chin, J.C. (1997) Peripheral blood lymphocyte subsets in fleece rot-resistant and -susceptible sheep. Australian Veterinary Journal 75, 421–423. McGuirk, B.J. and Atkins, K.D. (1984) Fleece rot in Merino sheep. I The heritability of fleece rot in unselected flocks of medium-wool Peppin Merinos. Australian Journal of Agricultural Research 35, 423–434. McGuirk, B.J. and Watts, J.E. (1983) Association between fleece, skin and body characters of sheep and susceptibility to fleece rot and body strike. In: Raadsma, H.W. (ed.) Sheep Blowfly and Flystrike in Sheep. Second National Symposium. New South Wales Department of Agriculture, Sydney, pp. 367–386. McGuirk, B.J., Atkins, D.A., Kowal, E. and Thornberry, K. (1978) Breeding for resistance to fleece rot and body strike – the Trangie programme. Wool Technology and Sheep Breeding 26(part iv), pp. 17–24. Mortimer, S.I., Atkins, K.D. and Raadsma, H.W. (1998) Responses to selection for resistance and susceptibility to fleece rot and body strike in Merino sheep. Proceedings of the 6th World Congress on Genetics applied to Livestock Production, p. 283. Mulder, K. and Colditz, I.G. (1993) Migratory responses of ovine neutrophils to inflammatory mediators in vitro and in vivo. Journal of Leukocyte Biology 53, 273–278. Nesa, M. (1994) The immune and inflammatory responses of Merino sheep bred for resistance

Host Resistance to Fleece Rot and Blowfly Strike

447

and susceptibility to fleece rot and flystrike: the Trangie R and S flocks. MSc Thesis, La Trobe University, Melbourne, Australia. O’Donnell, I.J., Green, P.E., Connell, J.A. and Hopkins, P.S. (1980) Immunoglobulin G antibodies to the antigens of Lucilia cuprina in the sera of fly-struck sheep. Australian Journal of Biological Sciences 33, 27–34. O’Meara, T.J. and Raadsma, H.W. (1995) Phenotypic and genetic indicators of resistance to ectopathogens. In: Gray, G.D., Woolaston, R.R. and Eaton, B.T. (eds) Breeding for Resistance to Infectious Diseases in Small Ruminants. ACIAR monograph no. 34, Canberra, pp. 187–218. O’Meara, T.J., Nesa, M., Raadsma, H.W., Saville, D.G. and Sandeman, R.M. (1992) Variation in skin inflammatory responses between sheep bred for resistance or susceptibility to fleece rot and blowfly strike. Research in Veterinary Science 52, 205–210. O’Meara, T.J., Nesa, M., Seaton, D.S. and Sandeman, R.M. (1995) A comparison of inflammatory exudates released from myiasis wounds on sheep bred for resistance or susceptibility to Lucilia cuprina. Veterinary Parasitology 56, 207–223. O’Meara, T.J., Nesa, M. and Sandeman, R.M. (1997) Antibody responses to Lucilia cuprina in sheep selected for resistance or susceptibility to L. cuprina. Parasite Immunology 19, 535–543. O’Sullivan, B.M., Hopkins, P.S. and Connell, J.A. (1984) The pathogenesis of flystrike in sheep. Animal Production in Australia 15, 171–181. Raadsma, H.W. (1989) Fleece rot and body strike in Merino sheep. III. Significance of fleece moisture following experimental induction of fleece rot. Australian Journal of Agricultural Research 40, 897–912. Raadsma, H.W. (1991) Fleece rot and body strike in Merino sheep. V. Heritability of liability to body strike in weaner sheep under flywave conditions. Australian Journal of Agricultural Research 42, 279–293. Raadsma, H.W. (1993) Fleece rot and body strike in Merino sheep. VI. Experimental evaluation of some physical fleece and body characteristics as indirect selection criteria for fleece rot. Australian Journal of Agricultural Research 44, 915–931. Raadsma, H.W. and Rogan, I.M. (1987) Genetic variation in resistance to blowfly strike. In: McGuirk, B.J. (ed.) Merino Improvement Programs in Australia. Proceedings of a National Symposium, Leura, NSW. Australian Wool Corporation, Melbourne, pp. 321–340. Raadsma, H.W. and Wilkinson, B.R. (1990) Fleece rot and body strike in Merino sheep. IV Experimental evaluation of traits related to greasy wool colour for indirect selection against fleece rot. Australian Journal of Agricultural Research 41, 139–153. Raadsma, H.W., Gilmour, A.R. and Paxton, W.J. (1989) Fleece rot and body strike in Merino sheep. II. Phenotypic and genetic variation in liability to fleece rot following experimental induction. Australian Journal of Agricultural Research 40, 207–220. Raadsma, H.W., Sandeman, R.M., Sasiak, A.B., Engwerda, C.R. and O’Meara, T.J. (1992) Genetic improvement in resistance to body strike in Merino sheep: where are we at with indirect selection. Proceedings of the Australian Association for Animal Breeding and Genetics 10, 143–146. Sandeman, R.M., Dowse, C.A. and Carnegie, P.R. (1985) Initial characterisation of the sheep immune response to infections of Lucilia cuprina. International Journal for Parasitology 15, 181–185. Sandeman, R.M., Collins, B.J. and Carnegie, P.R. (1987) A scanning electron microscope study of L. cuprina larvae and the development of blowfly strike in sheep. International Journal for Parasitology 17, 759–765. Seaton, D.S., O’Meara, T.J., Chandler, R.A. and Sandeman, R.M. (1992) The sheep antibody response to repeated infection with Lucilia cuprina. International Journal for Parasitology 22, 1169–1174.

448

I.G. Colditz and R.L. Tellam Seddon, H.R. and McGrath, T.T. (1929) Green colouration in wool. Agricultural Gazette of New South Wales 40, 206. Seddon, H.R., Belschner, H.G. and Mulhearn, C.R. (1931a) Observations on the susceptibility of sheep to blowfly attack. NSW Department of Agriculture Science Bulletin 37, 3–23. Seddon, H.R., Belschner, H.G. and Mulhearn, C.R. (1931b) The influence of breeding as a factor in the sheep blowfly problem. NSW Department of Agriculture Science Bulletin 37, 24–30. Stuart, T.P.A. (1894) On green-producing chromogenic microorganisms in wool. Journal and Proceedings of the Royal Society of New South Wales 28, 320–322. Sutherland, S.S., Ellis, T.M., Robertson, G.M. and Gregory, A.R. (1987) Serum and skin surface antibody responses in Merino sheep given three successive inoculations with Dermatophilus congolensis. Veterinary Microbiology 15, 209–218. Tellam, R.L., Smith, D., Kemp, D.H. and Willadsen, P. (1992) Vaccination against ticks. In: Yong, W.K. (ed.) Animal Parasite Control Utilizing Biotechnology. CRC Press, New York, pp. 303–331. Watson, D.L., Lea, J. and Burke, G.L. (1992) A method for collecting interstitial fluid from skin of sheep. Australian Veterinary Journal, 69, 14–15. Willadsen, P., Riding, G., McKenna, R.V., Kemp, D.H., Tellam, R.L., Nielsen, J.N., Lahnstein, J., Cobon, G. and Gough, J.M. (1989) Immunological control of a parasitic arthropod. Identification of a protective antigen from Boophilus microplus. Journal of Immunology 143, 1346–1351.

26

Host Resistance to Mastitis

K. PERSSON WALLER Swedish University of Agricultural Sciences, Faculty of Veterinary Medicine, Department of Obstetrics and Gynaecology, Uppsala, Sweden

Introduction Udder diseases, mainly mastitis induced by bacterial infections, are major problems for the dairy industry world wide, and the yearly costs are substantial (Philpot, 1984; Nilsson and Holmberg, 1996). Current programmes for control of bovine mastitis have improved the situation substantially, but the incidence of clinical and subclinical mastitis is still high. This indicates problems in the application of the control measures, or limitations in their effects. A need for new and innovative approaches for mastitis control especially concerning preventive measures is obvious. Lately, emphasis has been put on working towards a better understanding of the host resistance against udder infections, in order to find ways to increase the natural ability of the cow to resist, or defend herself, against udder infections. This paper will consist of a review of the latest knowledge in defence mechanisms of the mammary gland of ruminants. Special emphasis will be put on the distribution and function of various leucocyte populations, and the influence of different inflammatory mediators, such as cytokines, during the lactation cycle. Genetic influence on immunological functions will also be highlighted. Moreover, the influence of nutrition and stress on immunological functions and the defence against udder infections will be discussed, especially in relation to periods of immunosuppression, e.g. around calving, when the risk for mastitis is increased.

Specific and non-specific defence mechanisms of the mammary gland Mechanisms The defence mechanisms can be divided into non-specific (innate) and specific (adaptive/acquired) mechanisms. Leucocytes in udder secretions and udder tissues are important components of both categories. The non-specific response is essential during the © CAB International 2000. Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction (ed. P.B. Cronjé)

449

K. Persson Waller

450

early stages of infection and inflammation. Important considerations are anatomical features of the teat end, but also cell populations like macrophages, neutrophils, natural killer-like cells, and soluble factors like lactoferrin, lysozyme and complement. The specific response is mediated mainly by various antibodies (immunoglobulins) and lymphocytes recognizing specific determinants of the pathogen, but also by the effects of macrophages. The teat end and the teat canal form the first line of defence against infections. A well-functioning sphincter muscle, in combination with keratin lining the teat canal, ensures a secure mechanical closure of the canal (Craven and Williams, 1985). The keratin can trap bacteria entering the canal, and at each milking some of the keratin, with its possible content of bacteria, is desquamated and flushed out. The integrity of the teat skin of the teat end, especially around the teat orifice, is also important in order to minimize the numbers of bacteria present and thus reduce the risk for infection. The soluble factors involved in the host defence have direct antibacterial properties of varying efficacy, but also the ability to stimulate antibacterial functions of leucocytes, e.g. by opsonization. Moreover, combinations of certain soluble factors have synergistic effects. The concentrations of these substances are most often low in normal milk, but increase during involution of the udder and mastitis. Their importance during udder infections and mastitis is, to some extent, uncertain.

Distribution of different leucocyte populations The leucocytes, mainly neutrophils, macrophages and lymphocytes, have a key role in the defence of the udder. The proportions of each cell type in the mammary secretions vary with stage of lactation and health status in the gland (Concha, 1986; Paape et al., 1991). In normal milk and dry secretion, macrophages and lymphocytes are the predominating cell types while the numbers of neutrophils are low. However, during acute inflammation, the numbers of neutrophils increase dramatically. Chemotactic substances attract these cells from the blood to the site of inflammation. Prompt migration of sufficient numbers of neutrophils is essential to the outcome of an infection (Hill, 1994). Macrophages in tissues and udder secretions originate in blood monocytes and a selection for these cells, as expressed by CD14+ leucocytes, occurs in dry secretion compared with in milk (Persson Waller and Colditz, 1998). An increase in the number of lymphocytes occurs during mastitis, but to a lesser extent compared with the numbers of neutrophils. Most lymphocytes present in mammary secretions are either T-cells or B-cells (Craven and Williams, 1985; Concha, 1986). T lymphocytes can be further divided into ␣␤ T-cells consisting either of T-helper (CD4+) or T-suppressor/cytotoxic cells (CD8+), and ␥␦ T-cells (WC1+). A selective migration to, or retention of, CD8+ cells in favour of CD4+ cells occurs in the mammary gland (Taylor et al., 1994; Guiguen et al., 1996; Persson Waller and Colditz, 1998). The proportions of the different subpopulations in udder secretion and tissues, as well as expression of leucocyte antigens on mammary cells, differ with stage of lactation. In ewes, greater numbers of IL-2R+ and MHCII+ lymphocytes were observed in dry secretions compared with in milk (Persson

Host Resistance to Mastitis

451

Waller and Colditz, 1998). Memory cells of both B- and T-types are usually produced after an infection.

Leucocyte functions Phagocytosis and intracellular killing of bacteria are crucial functions of the neutrophils, which is illustrated by a higher prevalence of clinical mastitis when the cellular capacity for phagocytosis is low (Grommers et al., 1989; Hill, 1994). Their antimicrobial effects can be enhanced by opsonins, i.e. antibodies and complement (Craven and Williams, 1985; Paape et al., 1991). The functions of neutrophils may often be reduced in milk as they are affected by local conditions. However, newly recruited cells from the bloodstream are considered to have more efficient antibacterial properties (Herbelin et al., 1997), thus the importance of migration of sufficient numbers of neutrophils. Changes in adhesion molecules, shedding of L-selectin and up-regulation of CD18, occur in ovine neutrophils due to migration from blood into the udder (Persson Waller and Colditz, 1998). Macrophages are also phagocytic cells but their bactericidal capacity is lower than for neutrophils. The most important function of macrophages is probably their role as initiators of inflammatory processes (Adams and Hamilton, 1988). They recognize antigen, which they phagocytose, process and present to lymphocytes, which, in turn, become activated. Stimulated macrophages release inflammatory mediators important for inflammation and attraction of neutrophils (Adams and Hamilton, 1988). Also bovine mammary macrophages are capable of inducing proliferation of both blood and mammary lymphocytes (Concha and Holmberg, 1990). As mentioned earlier, lymphocytes can be divided into T-cells, involved in cellmediated immune response, and B-cells, which are precursors of antibody-producing plasma cells. Recent research has shown the importance of different lymphocyte subpopulations for the mammary immune response. CD8+ T-cells seem to have varying functions, being either of suppressor or cytotoxic type, depending on the stage of lactation (Shafer-Weaver and Sordillo, 1997). T-helper cells (CD4+) play an important role in the immune response by activation of other cells, like lymphocytes and macrophages (Tizard, 1996). The proportion of CD8+ lymphocytes in udder secretion is often larger than the proportion of CD4+ cells, which might have implications on the immune response of the mammary gland (Taylor et al., 1994; Guiguen et al., 1996; Persson Waller and Colditz, 1998). CD8+ cells have been shown to suppress immune function during udder infections (Park et al., 1993). The functions of ␥␦ T-cells in the udder are not certain but these cells are thought to be associated with protection of epithelial surfaces (Allison and Havran, 1991). Shafer-Weaver et al. (1996) found that the percentage of ␥␦ T lymphocytes decreased significantly in mammary parenchyma during times of increased susceptibility to infections. Antibody production is the primary role of B-cells. Antigen activation of cellular receptors results in a proliferation into antibody-secreting plasma cells which produce immunoglobulins directed against invading pathogens (Tizard, 1996). They also present antigen to T-helper lymphocytes, which produce cytokines inducing further proliferation and differentiation of B-cells.

452

K. Persson Waller

The inflammatory process and influence of inflammatory mediators (e.g. cytokines) Mastitis, or udder inflammation, is the host response to tissue damage, and is thereby proof of a functioning mammary defence. The inflammatory response is initiated through the production and/or the release of a complex cascade of inflammatory mediators resulting in a sequence of vascular and cellular events inducing the classical symptoms of inflammation. Among the inflammatory mediators responsible for these events are histamine, prostaglandins, leukotrienes and cytokines. To efficiently eliminate pathogens, the inflammatory reaction must be rapidly elicited and leucocytes, especially neutrophils, must quickly migrate in sufficient numbers to the udder. Cytokines are hormone-like soluble proteins and/or glucoproteins synthesized naturally by both immune and non-immune cells. They have important regulatory functions in most aspects of inflammation and immunity. Many cytokines have been identified, their roles during mastitis have been studied to some extent, and some cytokines have been used in research to treat or prevent mastitis in cows (Nickerson et al., 1989; Daley et al., 1993). Examples of some important cytokines are interferons (IFN), interleukins (IL) and tumour necrosis factors (TNF). The recruitment of neutrophils is an important part of the defence and cytokines are important for their migration. IFN-␥, IL-1␤ and TNF-␣ are cytokines known to stimulate the expression of adhesion molecules on circulating neutrophils and endothelial cells (Brown Cramer, 1992). Following adherence and the production of IL-8 by the endothelium, neutrophils then migrate into the extravascular space. Bovine mammary epithelial cells of lactating cows can produce IL-1 and IL-6, and these cells also have gene expression for multiple cytokines (Okada et al., 1997). Recent studies have suggested a role for cytokines in ruminant mastitis (Shuster et al., 1993; Shuster and Kehrli, 1995; Persson Waller et al., 1997a,b). During endotoxin-induced mastitis in sheep, TNF-␣, IL-8 and GM-CSF, but not IL-1␤ and IFN-␥, seem to be involved in leucocyte accumulation (Persson Waller et al., 1997b). The cytokine concentrations tend to increase just before, or together with, the increase in leucocyte accumulation. Cytokines were usually released during a short period and reached peak concentrations earlier than peak leucocyte infiltration (Persson Waller et al., 1997b). In contrast, in a similar study of bovine endotoxin-induced mastitis, high concentrations of IL-1 and IL-6 were found in milk but only small amounts of TNF-␣ and no IL-8 (Shuster et al., 1993; Shuster and Kehrli, 1995). However, IL-1 may not be an important mediator of endotoxin-induced inflammation in the bovine udder as intramammary infusion of an IL-1 receptor antagonist had no effect on the inflammatory response to endotoxin (Shuster and Kehrli, 1995; Persson Waller, 1997). TNF-␣, IL-8 and IL-1␤ were also detected in milk during ovine mastitis caused by experimental infection with Staphylococcus aureus or Escherichia coli (Persson Waller et al., 1997a). The kinetics of TNF-␣ and IL-8 influx were similar to that of leucocytes during both infections. However, IL-1␤ was detected mainly during S. aureus mastitis indicating differences in the inflammatory responses elicited by the organisms. TNF-␣ was also detected during bovine mastitis induced by E. coli (Sordillo and Peel, 1992). TNF-␣ is produced by mononuclear cells and this production is enhanced during the periparturient period which might explain the severe acute phase response of the mammary gland during this period (Sordillo et al., 1995). The presence of TNF-␣ and IL-6

Host Resistance to Mastitis

453

in sera appears to relate to the severity of clinical symptoms during coliform mastitis (Nakajima et al., 1997).

Genetic influences on immunological functions and the risk for udder infections General aspects Research on genetic resistance against udder infections and mastitis is important. However, heritability for clinical mastitis has been estimated to be low, on average 0.04 (e.g. Emanuelson et al., 1988; de Haas, 1998). Somatic cell count (SCC), as an indirect measurement of mastitis, has a higher heritability (on average 0.10), and might be a better selective tool than clinical mastitis (e.g. Emanuelson et al., 1988; de Haas, 1998). Moderate genetic correlation has been estimated between clinical mastitis and SCC. A combination of these parameters should therefore be considered for selection towards reduced susceptibility to mastitis. Kehrli and Shuster (1994) argued that too low SCC should be avoided as cows with very low SCC might be more susceptible to mastitis. However, Philipsson et al. (1995) reported a linear relationship between sires’ breeding values for clinical mastitis and for SCC. High milk production and various udder traits, such as udder suspensory ligament, fore udder attachment and udder depth, are associated with increased incidence of mastitis (Emanuelson, 1988; Uribe et al., 1995; Alexandersson, 1998). A similar relationship was observed between mastitis traits and milk protein production (de Haas, 1998). In the Nordic countries, udder traits, cell counts and clinical mastitis are included in the breeding values of bulls in order to improve disease resistance. In the Swedish selection programme, diseases other than mastitis are also included in the breeding values. A combined index of breeding values for mastitis and SCC was introduced in the beginning of this decade (Koenen et al., 1994). Other countries have recently started to include mastitis selection parameters in their breeding programmes.

Genetics and the immune response Recently, the genetic influence on immunological parameters has been studied. These are aimed towards the possibility of selecting cows with improved immune response, without negative effects on milk productivity. Differences in proportions of leucocyte subpopulations may have an important role in the probability of an individual cow developing mastitis, as the proportions of polymorphonuclear leucocytes, macrophages, and CD4+ or CD8+ T-lymphocytes differ significantly between cows (Leitner et al., 1998). Kelm et al. (1997) used different measures of mastitis, and a number of immunological assays and molecular markers. The molecular markers accounted for up to 40% of the variation in estimated breeding value (EBV) for measures of mastitis. In cows with a low EBV for somatic cell score, the neutrophils functioned more effectively at maximal immunosuppression, the serum immunoglobulin concentration was lower, and the numbers of circulating mononuclear

K. Persson Waller

454

cells were higher. Significant differences in sire progeny groups for various neutrophil assays have also been observed (Kehrli et al., 1991), and results of Fitzpatrick et al. (1998) indicated that immunological assays may be useful in identifying bulls whose progeny would be associated with a higher resistance to intramammary infections with S. aureus. A selection programme for animals resistant to immunosuppression and disease around parturition would be especially desirable. Genetic variability in certain immune functions, i.e. total numbers of neutrophils, neutrophil chemokinesis, assays of neutrophil respiratory burst associated with phagocytosis, and in serum concentrations of immunoglobulins and hemolytic complement activity was observed in periparturient dairy cows (Detilleux et al., 1994). Moreover, Mallard et al. (1998) showed that cows at peripartum may be categorized as high or low responders to immunizations and that the heritability for specific antibody responses was moderate to high. These findings may be used to identify animals with high resistance to disease. The importance of adhesion molecules, essential for leucocyte migration, has recently been highlighted. Genetic leucocyte adhesion deficiencies leading to chronic, or even fatal, infections have been observed in several animal species including cattle (bovine leucocyte adhesion deficiency) (Kehrli et al., 1992). All species affected show signs of chronic and recurrent infections due to a deficiency in the chemotactic and phagocytic properties of leucocytes, particularly neutrophils.

Influence of stress on immune functions, and periods of immunosuppression Influence of stress Stress has a negative effect on the immune response and can make both adults and young animals less resistant to infectious diseases. This is probably mainly due to increased levels of endogenous corticosteroids which can affect both the functions and numbers of leucocytes, and thus increase the host’s susceptibilty to infections (Guidry et al., 1976). Treatment of cows with synthetic glucocorticoids, like dexamethasone, have a negative influence on immune functions, e.g. affecting the numbers and proportions of leucocytes in blood due to down-regulation of adhesion molecules on blood neutrophils (Burton and Kehrli, 1995; Burton et al., 1995). Such changes were associated with leucocytosis and increased shedding of S. aureus in subclinically infected mammary glands, which reinforces the potential risk of treating infected animals with potent synthetic glucocorticoids such as dexamethasone.

Periods of immunosuppression As a result of depressed immune functions, susceptibility to infectious diseases, such as mastitis, is associated with the peripartum period (Sordillo et al., 1997). High blood levels of glucocorticoids are present during this time. Examples of stress factors during this period are parturition, onset of lactation, and changes in feeding and management

Host Resistance to Mastitis

455

regimes. Impaired functions of a number of immunological parameters have been observed in peripartal dairy cows. Amongst these are, for example, changes in the white blood cell count and in blood and mammary neutrophil and lymphocyte functions. The cell functions decrease during the weeks before calving, reach their lowest levels at calving, and return to initial preparturient levels one to three weeks after calving (e.g. Saad et al., 1989; Detilleux et al., 1994; Shafer-Weaver et al., 1996; Lee and Kehrli, 1998). A reduced function of the blood cells diminishes their ability to kill bacteria and to stimulate the immune defence. A significant negative correlation between severity of E. coli mastitis and the number of circulating mature neutrophils in blood immediately before infection has been demonstrated (Dosogne et al., 1997). Preferential trafficking of T-cells of the suppressor type rather than of the cytotoxic type into the mammary gland tissues around calving, compared with in mid-lactation, may be another factor responsible for the lowered local immune responsiveness during this period (Sordillo et al., 1995). Moreover, Sordillo et al. (1991) reported that the levels of the cytokines IL-2 and IFN-g in mammary gland secretions decrease before calving. Both cytokines are essential for several important immune functions, and lowered concentrations may contribute to the increased susceptibility to infections.

Influence of some aspects of nutrition on immunological functions and the risk for udder infections General aspects Adequate nutrition is essential for a good immune response and thus for overall health and the ability of the animal to resist diseases. The balance between protein and energy, the feeding routines and eating time are factors influencing animal health by potentially increasing the risk for metabolic disturbances. Differences in feeding strategy were observed between farms with a high or low incidence of clinical mastitis (Hallén Sandgren, 1998). Also, the hygienic quality of the ration is essential. At the onset of lactation dairy cows are metabolically stressed due to negative energy balance leading to the mobilization of considerable amounts of tissue reserves. This could also influence the immune functions as increased levels of ␤-hydroxybutyrate and other ketone bodies can have negative effects on cellular functions (e.g. Klucinski et al., 1988; Hoeben et al., 1997). Balanced feeding during the dry period is a prerequisite for good feed consumption after calving thus avoiding metabolic stress and increased risk for disease during the early lactation period. Dairy cows often suffer from hepatic lipidosis during early lactation, increasing the risk for metabolic diseases. This may further decrease the immune response, as a high fat content of the liver is associated with decreased function of granulocytes (Zerbe et al., 1998). The amino acid glutamine has unique metabolic functions in dairy cows. Demands for glutamine increase with increased metabolic stress at the onset of lactation (Lacey and Wilmore, 1990; Hall et al., 1996). Substantial amounts of glutamine are metabolized in the udder where it is used for synthesis of milk protein, as a source of energy and as substrate for cell proliferation (Meijer et al., 1993). Glutamine is a key substance also for the immune defence, e.g. for the proliferation of lymphocytes and macrophages (Rohde et al., 1996).

K. Persson Waller

456

In a recent study, K. Holtenius, K. Persson Waller, B. Essén-Gustavsson and P. Holtenius (unpublished data) investigated the connection between metabolic stress around calving/early lactation and the incidence of clinical mastitis in high-producing Swedish cows. All herds had low bulk milk SCC, but had either a low (LT) or a high (HT) treatment incidence of mastitis. The results indicate that cows in the HT group mobilize their tissue reserves of fat and protein to a greater extent, during the weeks after calving, than cows in the LT group. The concentrations of glutamine decreased during late pregnancy and early lactation, and was lower (P = 0.09) in cows from HT herds during early lactation. Moreover, the blood concentrations of tryptophane were lower (P < 0.01) and the free fatty acids were higher (P = 0.04) in HT herds during early lactation. These differences could be due to increased metabolic stress in these herds. The results also indicate that the metabolic changes started earlier than 3 weeks before calving. Preliminary results indicate that there were differences in feeding strategies between the groups: the HT group was fed more concentrates and less roughage (C. Hallén Sandgren, L. Elander and U. Emanuelson, unpublished data). There were probably other management differences as well. The results of the blood leucocyte counts did also indicate a higher stress level in the HT group, which possibly affected the immune defence of these animals. The numbers of eosinophils were lower (P = 0.08) in the HT group, while the numbers of neutrophils were higher (P = 0.06) in this group. Lower numbers of eosinophils and higher numbers of neutrophils can be a result of stress-induced release of cortisol. This is associated with a negative effect on the adhesion and migration ability of the neutrophils. An alternative explanation to the higher numbers of neutrophils is an increased recruitment of cells from the bone marrow due to some inflammatory stimulus in the body.

Micronutrients A balanced supply of micronutrients (vitamins and trace elements) is essential during periods of immune suppression. Deficiencies in selenium (Se), vitamin E, vitamin A, copper (Cu) and zinc (Zn) have been associated with a negative influence on the immune response in association with mastitis (Reddy and Frey, 1990; Harmon and Torre, 1994; Smith et al., 1997). Se deficiency is a problem in areas where Se concentrations in soils and pastures are low. Se is an important component of the enzyme glutathione peroxidase, which is essential for the protection of cells and tissues from auto-oxidative damage from production of oxygen radicals (Reddy and Frey, 1990). Se deficiency results in reduced neutrophil migration into the udder and impaired intracellular killing of bacteria (Erskine, 1993; Hogan et al., 1993). Supplementation of Se may improve udder health by reducing the severity and duration of cases of clinical mastitis (Erskine, 1993). Vitamin E is important for both cellular and humoral immune functions. This may be elicited through its effect on cell membrane stability and regulatory role in biosynthesis of various inflammatory mediators (Reddy and Frey, 1990; Smith et al., 1997). Dietary supplementation of vitamin E is of value, especially as the concentration of this vitamin in fodder decreases with age and length of storage. Moreover, the serum concentration of vitamin E drops around calving, a period of increased suscept-

Host Resistance to Mastitis

457

ibility to disease (Smith et al., 1997). Supplementation with vitamin E increased intracellular killing of bacteria by neutrophils, and reduced the numbers of new intramammary infections (Smith et al., 1997). Dietary supplementation with vitamin E should be considered as a preventive measure in the control of mastitis during the peripartal period. Vitamin A and the precursor ␤-carotene are also important for mucosal integrity and stability. Both substances have stimulatory effects on immune cell populations and have been correlated with an increased resistance to disease (Sordillo et al., 1997). Deficiencies were related with severity of mastitis, and there is a negative correlation between plasma vitamin A and somatic cell counts in milk (Johnston and Chew, 1984). Plasma vitamin A and ␤-carotene concentrations decrease during the periparturient period contributing towards increased susceptibility to new udder infections (Johnston and Chew, 1984). Limited information is available on the importance of Cu and Zn in minimizing the risk for mastitis in dairy cows. However, Cu and Zn have important biological functions, such as being parts of enzyme systems with antioxidative properties. These systems protect cells and tissues from detrimental effects of oxidative substances released at phagocytosis and killing of bacteria by white blood cells (Reddy and Frey, 1990). A Cu deficiency has also been reported to decrease the antibacterial effects of the immune defence (Reddy and Frey, 1990; Harmon and Torre, 1994). The addition of Cu in the feed gave a considerable decrease in the numbers of infected udder quarters at calving compared with untreated controls (Harmon and Torre, 1994). Moreover, the cell count in milk tended to be lower in the group receiving Cu supplementation. Zn is important for the integrity of the skin, the first barrier against infections, and has also importance for immune functions. Zn deficiency can cause degeneration of lymphoid tissues and have a negative influence on immune cells (Reddy and Frey, 1990; Harmon and Torre, 1994). Few studies have been made to study the relationship between Zn and mastitis. However, deficiency in Zn can predispose cows to secondary infections, which can be reversed by supplementation of Zn (Reddy and Frey, 1990; Harmon and Torre, 1994). Problems associated with Zn deficiency can be exacerbated by having a feeding regime containing high amounts of calcium during early lactation. Plasma Zn concentrations decrease markedly at parturition which may be connected with the immune suppression observed during this period (Goff and Stabel, 1990).

Conclusions Proper immune functions are essential for the defence against udder infections. Detailed knowledge about the immune response and important defence factors are essential in order to find new ways for the prevention and treatment of udder infections leading to mastitis. Work should be concentrated on ways of minimizing the negative influence on immune functions and/or ways of stimulating these functions, especially during periods of immune suppression such as around calving. Possibilities of stimulating the non-specific immune response, for example through the use of cytokines and other immunomodulatory substances, should be comprehensively investigated. As mentioned above, it is important to identify risk factors, which negatively influence the defence mechanisms of the udder. The importance of management and

K. Persson Waller

458

adequate nutrition are some important factors to consider. Provision of suitable dietary supplies of vitamins and trace elements is one important step to ensure a good mammary defence and prevent mastitis. The possibility to find markers for genetic selection of individuals with a well-developed immune system should also be further evaluated.

References Adams, D.O. and Hamilton, T.A. (1988) Phagocytic cells: Cytotoxic activities of macrophages. In: Gallin, J.I., Goldstein, I.M. and Snyderman, R. (eds) Inflammation: Basic Principles and Clinical Correlates. Raven Press, New York. Alexandersson, G. (1998) Genetiska Samband mellan Exteriöra Juveregenskaper och Klinisk Mastit hos Förstakalvande SRB-kor. Publication 195, Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, Sweden. Allison, J.P. and Havran, W.L. (1991) The immunobiology of T cells with invariant gamma delta antigen receptors. Annual Review of Immunology 9, 679–705. Brown Cramer, E. (1992) Cell biology of phagocyte migration from the bone marrow, out of the bloodstream, and across organ epithelia. In: Gallin, J.I., Goldstein, I.M. and Snyderman, R. (eds) Inflammation: Basic Principles and Clinical Correlates. Raven Press, New York, pp. 341–352. Burton, J.L. and Kehrli, M.E., Jr (1995) Regulation of neutrophil adhesion molecules and shedding of Staphylococcus aureus in milk of cortisol- and dexamethasone-treated cows. American Journal of Veterinary Research 56, 997–1006. Burton, J.L., Kehrli, M.E., Jr, Kapil, S. and Horst, R.L. (1995) Regulation of L-selectin and CD18 on bovine neutrophils by glucocorticoids: effects of cortisol and dexamethasone. Journal of Leukocyte Biology 57, 317–325. Concha, C. (1986) Cell types and their immunological functions in bovine mammary tissues and secretions – A review of the literature. Nordic Veterinary Medicine 38, 257–272. Concha, C. and Holmberg, O. (1990) Ability of bovine mammary macrophages to enhance proliferation of autologous blood and mammary secretion lymphocytes. Journal of Dairy Research 57, 7–16. Craven, N. and Williams, M.R. (1985) Defences of the bovine mammary gland against infection and prospects for their enhancement. Veterinary Immunology and Immunopathology 10, 71–127. Daley, M.J., Williams, T., Coyle, P., Furda, G., Dougherty, R. and Hayes, P.W. (1993) Prevention and treatment of Staphylococcus aureus infections with recombinant cytokines. Cytokine 5, 276–284. de Haas, Y. (1998) Genetic Parameters of Mastitis Traits and Protein Production in the First Two Lactations of Swedish Red and White Cows. Publication no. 17. Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, Uppsala, Sweden. Detilleux, J.C., Koehler, K.J., Freeman, A.E., Kehrli, M.E., Jr and Kelley, D.H. (1994) Immunological parameters of periparturient Holstein cattle: Genetic variation. Journal of Dairy Science 77, 2640–2650. Dosogne, H., Burvenich, C., van Werven, T., Roets, E., Noordhuizen-Stassen, E.N. and Goddeeris, B. (1997) Increased surface expression of CD11b receptors on polymorphonuclear leukocytes is not sufficient to sustain phagocytosis during Escherichia coli mastitis in early postpartum dairy cows. Veterinary Immunology and Immunopathology 60, 47–59. Emanuelson, U., Danell, B. and Philipsson, J. (1988) Genetic parameters for clinical mastitis, somatic cell counts, and milk production estimated by multiple-trait restricted maximum likelihood. Journal of Dairy Science 71, 467–476.

Host Resistance to Mastitis

459

Erskine, R.J. (1993) Nutrition and mastitis. Veterinary Clinics of North America Food Animal Practice 9, 551–559. Fitzpatrick, J.L., Logan, K.E., Stear, M.J., McGuirk, B. and Platt, D.J. (1998) Immunological responses of Holstein/Friesian cattle to Staphylococcus aureus. In: Proceedings of the 5th International Veterinary Immunology Symposium, November 8–13, Punjab Agricultural University, Ludhiana, India, p. 117. Goff, J.P. and Stabel, J.R. (1990) Decreased plasma retinol, ␣-tocopherol, and zinc concentration during the periparturient period: Effect of milk fever. Journal of Dairy Science 73, 3195–3199. Grommers, F.J., van de Geer, D., van der Vliet, H., Henricks, P.A. and Nijkamp, F.P. (1989) Polymorphonuclear leucocyte function: Relationship between induced migration into the bovine mammary gland and in vitro cell activity. Veterinary Immunology and Immunopathology 23, 75–83. Guidry, A.J., Paape, M.J. and Pearson, R.E. (1976) Effects of parturition and lactation on blood and milk cell concentrations, corticosteroids, and neutrophil phagocytosis in the cow. American Journal of Veterinary Research, 37, 1195–1200. Guiguen, F., Greenland, T., Pardo, E., Panaye, G. and Mornex, J.F. (1996) Flow cytometric analysis of goat milk lymphocytes: Subpopulations and adhesion molecule expression. Veterinary Immunology and Immunopathology 53, 173–178. Hall, J.C., Heel, K. and McCauley, R. (1996) Glutamine: A review. British Journal of Surgery 83, 305–312. Hallén Sandgren, C. (1998) Field experiences of mastitis risk factors with special emphasis on heifer management and replacement. In: Proceedings of the Nordic Seminar Regarding Future Aspects of Mastitis Prevention, January 21–22, Aarhus, Denmark. Harmon, R.J. and Torre, R.M. (1994) Copper and zinc: do they influence mastitis? In: Proceedings of the 33rd Annual Meeting of the National Mastitis Council, Orlando, Florida, National Mastitis Council, Inc., Arlington, Virginia, USA, p. 54. Herbelin, C., Poutrel, B., Gilbert, F.B. and Rainard, P. (1997) Immune recruitment and bactericidal activity of neutrophils in milk of cows vaccinated with staphylococcal alpha-toxin. Journal of Dairy Science 80, 2025–2034. Hill, A.W. (1994) Escherichia coli mastitis. In: Gyles, C.L. (ed.) Escherichia coli in Domestic Animals and Humans. CAB International, Wallingford, UK, pp. 117–133. Hoeben, D., Heyneman, R. and Burvenich, C. (1997) Elevated levels of beta-hydroxybutyric acid in periparturient cows and in vitro effect on respiratory burst activity of blood neutrophils. Veterinary Immunology and Immunopathology, 58, 165–170. Hogan, J.S., Weiss, W.P. and Smith, K.L. (1993) Role of vitamin E and selenium in host defense against mastitis. Journal of Dairy Science 76, 2795–2803. Johnston, L.A. and Chew, B.P. (1984) Peripartum changes of plasma and milk vitamin A and ␤carotene among dairy cows with or without mastitis. Journal of Dairy Science, 67, 1832–1840. Kehrli, M.E. and Shuster, D.E. (1994) Factors affecting milk somatic cells and their role in health of the bovine mammary gland. Journal of Dairy Science 77, 619–627. Kehrli, M.E., Jr, Weigel, K.A., Freeman, A.E., Thurston, J.R. and Kelley, D.H. (1991) Bovine sire effects on daughters’ in vitro blood neutrophil functions, lymphocyte blastogenesis, serum complement and conglutinin levels. Veterinary Immunology and Immunopathology 27, 303–319. Kehrli, M.E., Jr, Shuster, D.E. and Ackermann, M.R. (1992) Leukocyte adhesion deficiency among Holstein cattle. Cornell Veterinarian 82, 103–109. Kelm, S.C., Detilleux, J.C., Freeman, A.E., Kehrli, M.E., Jr, Dietz, A.B., Fox, L.K., Butler, J.E., Kasckovics, I. and Kelley, D.H. (1997) Genetic association between parameters of innate immunity and measures of mastitis in periparturient Holstein cattle. Journal of Dairy Science 80, 1767–1775.

460

K. Persson Waller Klucinski, W., Degorski, A., Miernik-Degorska, E., Targowski, S. and Winnicka, A. (1988) Effect of ketone bodies on the phagocytic activity of bovine milk macrophages and polymorphonuclear leucocytes. Zentralblatt für Veterinarmedizin A, 35, 632–639. Koenen, E.P.C., Berglund, B., Philipsson, J. and Groen, A.F. (1994) Genetic parameters of fertility disorders and mastitis in the Swedish Friesian breed. Acta Agricultura Scandinavica 44, 202–207. Lacey, J.M. and Wilmore, D.W. (1990) Is glutamine a conditionally essential amino acid? Nutrition Reviews 48, 297–309. Lee, E.K. and Kehrli, M.E., Jr (1998) Expression of adhesion molecules on neutrophils of periparturient cows and neonatal calves. American Journal of Veterinary Research 59, 37–43. Leitner, G., Chaffer, M., Winkler, M., Glickman, A., Weisblit, L., Krifucks, O., Ezra, E. and Saran, A. (1998) The determination of milk leukocyte population in heifers free of udder infection as a tool for the genetic improvement of dairy cows. In: Proceedings of the 5th International Veterinary Immunology Symposium, November 8–13, Punjab Agricultural University, Ludhiana, India, p. 119. Mallard, B.A., Dekkers, J.C., Ireland, M.J., Leslie, K.E., Sharif, S., Van Kampen, C.L., Wagter, L. and Wilkie, B.N. (1998) Alteration in immune responsiveness during the peripartum period and its ramification on dairy cow and calf health. Journal of Dairy Science 81, 585–595. Meijer, G.A., van der Meulen, J. and van Vuuren, A.M. (1993) Glutamine is a potentially limiting amino acid for milk production in dairy cows: A hypothesis. Metabolism 42, 358–364. Nakajima, Y., Mikami, O., Yoshioka, M., Motoi, Y., Ito, T., Ishikawa, Y., Fuse, M., Nakano, K. and Yasukawa, K. (1997) Elevated levels of tumor necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6) activities in the sera and milk of cows with naturally occurring coliform mastitis. Research in Veterinary Science 62, 297–298. Nickerson, S.C., Owens, W.E. and Watts, J.L. (1989) Effects of recombinant granulocyte colony-stimulating factor on Staphylococcus aureus mastitis in lactating dairy cows. Journal of Dairy Science 72, 3286–3294. Nilsson, L. and Holmberg, O. (1996) Research projects in mastitis. In: National Veterinary Institute, Research and Development for Animal Health. Uppsala, Sweden. Okada, H., Ito, T., Ohtsuka, H., Kirisawa, R., Iwai, H., Yamashita, K., Yoshino, T. and Rosol, T.J. (1997) Detection of interleukin-1 and interleukin-6 on cryopreserved bovine mammary epithelial cells in vitro. Journal of Veterinary Medical Science 59, 503–507. Paape, M.J., Guidry, A.J., Jain, N.C. and Miller, R.H. (1991) Leukocytic defense mechanisms in the udder. Flemish Veterinary Journal 62 (Suppl. 1), 95–109. Park, Y.H., Fox, L.K., Hamilton, M.J. and Davis, W.C. (1993) Suppression of proliferative response of BoCD4+ T lymphocytes by activated BoCD8+ T lymphocytes in the mammary gland of cows with Staphylococcus aureus mastitis. Veterinary Immunology and Immunopathology 36, 137–151. Persson Waller, K. (1997) Modulation of endotoxin-induced inflammation in the bovine teat using antagonists/inhibitors to leukotrienes, platelet activating factor and interleukin 1␤. Veterinary Immunology and Immunopathology 57, 239–251. Persson Waller, K. and Colditz, I.G. (1998) Expression of surface antigens on blood and mammary leukocytes in lactating and dry ewes. Veterinary Immunology and Immunopathology 62, 237–278. Persson Waller, K., Colditz, I.G. and Seow, H.-F. (1997a) Accumulation of leucocytes and cytokines in the lactating ovine udder during mastitis due to Staphylococcus aureus and Escherichia coli. Research in Veterinary Science 62, 63–66. Persson Waller, K., Colditz, I.G., Flapper, P. and Seow, H.-F. (1997b) Leukocyte and cytokine accumulation in the ovine teat and udder during endotoxin-induced inflammation. Veterinary Research Communications 21, 101–115.

Host Resistance to Mastitis

461

Philipsson, J., Ral, G. and Berglund, B. (1995) Somatic cell count as a selection criterion for mastitis resistance in dairy cattle. Acta Agricultura Scandinavica 41, 195–200. Philpot, W.N. (1984) Economics of mastitis control. Veterinary Clinics of North America 6, 233–245. Reddy, P.G. and Frey, R.A. (1990) Nutritional modulation of immunity in domestic food animals. Advances in Veterinary Science and Comparative Medicine 35, 255–281. Rohde, T., Mclean, D.A. and Klarlund-Petersen, B. (1996) Glutamine, lymphocyte proliferation and cytokine production. Scandinavian Journal of Immunology 44, 648–650. Saad, A.M., Concha, C. and Astrom, G. (1989) Alterations in neutrophil phagocytosis and lymphocyte blastogenesis in dairy cows around parturition. Zentralblatt für Veterinarmedizin Reihe B, 36, 337–345. Shafer-Weaver, K.A. and Sordillo, L.M. (1997) Bovine CD8+ suppressor lymphocytes alter immune responsiveness during the postpartum period. Veterinary Immunology and Immunopathology 56, 53–64. Shafer-Weaver, K.A., Pighetti, G.M. and Sordillo, L.M. (1996) Diminished mammary gland lymphocyte functions parallel shifts in trafficking patterns during the postpartum period. Proceedings of the Society for Experimental Biology and Medicine 212, 271–279. Shuster, D.E. and Kehrli, M.E., Jr (1995) Administration of recombinant human interleukin 1 receptor antagonist during endotoxin-induced mastitis in cows. American Journal of Veterinary Research 56, 313–320. Shuster, D.E., Kehrli, M.E. and Stevens, M.G. (1993) Cytokine production during endotoxininduced mastitis in lactating dairy cows. American Journal of Veterinary Research 54, 80–85. Smith, K.L., Hogan, J.S. and Weiss, W.P. (1997) Dietary vitamin E and selenium affect mastitis and milk quality. Journal of Animal Science 75, 1659–1665. Sordillo, L.M. and Peel, J.E. (1992) Effect of interferon-␥ on the production of tumor necrosis factor during acute Escherichia coli mastitis. Journal of Dairy Science 75, 2119–2125. Sordillo, L.M., Redmond, M.J., Campos, M., Warren, L. and Babiuk, L.A. (1991) Cytokine activity in bovine mammary gland secretions during the periparturient period. Canadian Journal of Veterinary Research 55, 298–301. Sordillo, L.M., Pighetti, G.M. and Davis, M.R. (1995) Enhanced production of bovine tumor necrosis factor-alpha during the periparturient period. Veterinary Immunology and Immunopathology 49, 263–270. Sordillo, L.M., Shafer-Weaver, K. and DeRosa, D. (1997) Immunobiology of the mammary gland. Journal of Dairy Science 80, 1851–1865. Taylor, B.C., Dellinger, J.D., Cullor, J.S. and Stott, J. (1994) Bovine milk lymphocytes display the phenotype of memory T cells and are predominantly CD8+. Cellular Immunology 156, 245–253. Tizard, I.R. (1996) Veterinary Immunology: An Introduction, 5th edn. W.B. Saunders Company, Philadelphia, USA, 531 pp. Uribe, H.A., Kennedy, B.W., Martin, S.W. and Kelton, D.F. (1995) Genetic parameters for common health disorders of Holstein cows. Journal of Dairy Science 78, 421–430. Zerbe, H., Schneider, N., Obadnik, C., Schuberth, H.-J., Wensing, T., Kruip, T., Grunert, E. and Leibold, W. (1998) The negative energy balance reduces the neutrophil defense capacity in cows. In: Proceedings of the 5th International Veterinary Immunology Symposium, November 8–13. Punjab Agricultural University, Ludhiana, India, p. 185.

Index

Figures in bold indicate major references. Figures in italic refer to diagrams, photographs and tables. abdominal discomfort 9 abortion 308 acid labile subunit (ALS) 191 acid–base balance 161–162 acidosis 161, 161 ADH (antidiuretic hormone) 45, 46 adhesion leucocyte 454 microbial 79–92 adipocytes 171, 173–174, 173, 176 adipose tissue 171–181 dairy cattle 357, 358–360, 358 hormone production 9 interactions with immune system 177–178 mammalian evolution 171–172 sheep 206–208, 208, 213–215 substances secreted 173–174, 173 adipsin 177 afferent pathways 8–11 allometric relationships 27, 27 amino acids fermenting bacteria 100–102 metabolism absorption 143 glutamine 161–162, 161 individual amino acids 158–160, 159, 160

liver N balance 157 plasma versus blood transport 157–158 requirements of gut tissues 141–143, 142 wool growth see wool growth ammonia assimilation 106 ammonia metabolism 151–152 bacteria 100–101, 106 control of ureagenesis 155–157 ornithine cycle 152–154, 153 capacity and substrate priority 154–155 amygdala 44–47, 46 amylolytic microbes 32 anabolic metabolites 155 animal factors 26–27 animal–microbe relationships 61–62 antibiotic resistance 390–391, 390 antibody production 451 Archaea 64 ATP 31, 35 autocrine effects 174–176 autolysis 108–109

B-cells 450, 451 bacteria 63 463

Index

464 amino-acid-fermenting 100–101 cellulolytic 79, 82 culturable number 101 cyanobacteria 110 Gram-negative 85 Gram-positive 68, 85, 100 mucosal 71 non-cellulolytic 82 obligate peptide 100–101 phylogenetic placement 68 proteolytic 101–102 ruminal 87, 108–110 bacterial glycoside hydrolases 83–84, 86–87 bacteriophages 103, 391–392 Bacteroideacea 101 Bacteroides distasonis 71 Bacteroides fragilis 71 barley 7 barley straw 13 Barnacle geese 172 biotechnology 399–402, 400 blackbrush 8 blood flow 132–133, 134 hepatic 150 blood pressure 46 blood transport, amino acids 157–158 blowfly strike 437 immunity to 440–441 resistance 438–440 vaccines 441–444 body condition score (BCS) 354–356, 356 body heat 295–297, 296 body temperature birth-related changes 303–305, 304, 307–308 elevated 47 feto-maternal relationship 297–299, 298, 299, 300, 302 homeostasis 295–297, 296 body tissues 9 body weight 198, 206 bone 188 brain 45, 211–212 photoperiodic changes 211–212 brainstem control of gastric motility 47–48 brain stimulation 48, 49 interior sites 50–51, 50, 53 surface sites 49–50 hypermotility 53–54 reticuloruminal motility 54–56, 55

rumination 53–54 control of salivation 42–44, 43 amygdala see amygdala frontal lobe 44 branch-chain amino acids (BCAA) 158 buoyancy 28–29 Butyrivibrio fibrisolvens 65, 68, 101, 102, 394–397, 396

calving 356, 455, 456 cAMP 89 carbon metabolism 108, 109 casein 5, 6 catalytic activity 91 catalytic domain (CD) 86, 87 catbolism 154, 198–199 catecholamines 206–207, 358 cattle see dairy cattle CBDs (cellulose binding domain) 80, 87, 88, 90 cDNA 179 CelA CBD 87 cellobiosidases 86 cellular responses (resistance) 427 cellulose-binding polypeptides 85 cellulose-binding proteins (CBPs) 80, 82, 85 cellulosome paradigm 80, 81, 86 chromatographic enrichment 105 chylomicron 175 CipA 80, 82 clostridia 80, 100, 101 Clostridium coccoides 71 Clostridium proteoclasticum 102 Clostridium thermocellum 80, 81 CNS (central nervous system) afferent pathways 8–9 body tissues 9 dietary choice 3, 4, 8 sensations from receptors 9–10 ‘discomfort’ signal 14–16, 15 CoA-esters 136 comfort level optimum 12–14 single food 11–12 comminution 28 compartment model (digestion) 25 competitive PCR (cPCR) 101–102 complement system 177 concentrates 7 conditioned preferences (food) 5

Index

465 conditioning stimuli 8–11 copper (Cu) 457 Cretaceous period 62 cyanobacteria 110 cyclical intake 11–12 cysteine proteinase 105 cytokines 177–178, 179, 188–189, 452

daily metabolizable energy (ME) 5 dairy cattle 353–354 adipose tissue 172, 358–360 body fat reserves 354–357, 356, 365 genetic merit 357, 358 lactating 364 body protein reserves 360–361 amino acid metabolism 362–363 muscle tissue metabolism 361–362 food choice 10 immunosuppression 454–455 nutrition 455–457 mastitis see mastitis metabolism 357–358, 357, 358 nutrient interconversions 363–366, 364, 365, 366 portal vein 132 rumen function 7 Dasytricha ruminantium 67 dehydration 47 DGGE (denaturing gradient gel electrophoresis) 69–70 diabetogenic effects 197 diet, change of 13–14 dietary choice 3–17 differential display RT-PCR 107–108 digesta kinetic models 21 animal factors 26–27 digesta clearance 22, 24–25, 25 digesta load 30–31 digestion kinetics 26 passage kinetics 27–29 performance and sensitivity 29–30 substrate fractionation 25–26 digestion 382 microbial 62–63 dinosaurs 62 discomfort 14–16, 15 disease resistance 382–383

DNA free 393 gut ecology 63–64, 65 PCR amplification 69 dockerins 82 dorsal vagal motor nucleus 48–49 dry matter (DM) 12 duodenum 28 dye-dilution procedures 132–133

ecosystems, microbial 61–72 effector phenomena 426 electrical stimulation (brain) 46, 48, 53–54 electrolyte movement 429 electrophoresis 69–70 endocrine system 187, 376–378 genotype differences 415–418 glucose transport 140 endogenous corticosteroids 454 endoglucanase activity 81, 85 endosymbiotic microbes 63 endotoxin 199–200 energy supply 6, 9 Enterobacteriaceae 63 Entodinium simplex 67 Entodinium species 67 enzymatic hydrolysis 79–92 enzyme supplementation 90 eosinophils 456 epicardial adipose tissue 176–177 Epidinium caudatum 67 epithelial sloughing 429 Escherichia coli 71, 85, 452, 455 estimated breeding value (EBV) 453–454 ethyl methane sulphonate (EMS) 104 Eubacterium 101, 102 Eubacterium rectale 71 Eucarya 64

fat cell size 356 fat corrected milk (FCM) 356 fatty acids 175–176 feeding neophobia 4 fermentation 24–25, 25 bacterial 62, 63, 66 foregut 62–63 hindgut 63 ruminal 135 fetal programming 241–243

Index

466 fetus abortion 308 growth 192–193 thermal physiology birth-related changes 303–305, 304, 307–308 feto-maternal relationship body temperatures 297–299, 298, 299, 300, 302 fetal hyperthermia 299–300 maternal hyperthermia 301–303 maternal hypothermia 303 homeostasis 295–297, 296 thermoregulatory mechanisms 305–307, 306 fibre 7, 15, 86, 104 Fibrobacter 65, 66 Fibrobacter succinogenes 66, 69–70, 79, 82, 89, 108, 398 fibrolytic enzymes 90 fibrolytic microbes 32 flavours 4–5 flow rates microbes 32 nutrients 28 rumen liquid 32 fluorescent probes 71 follicles wool 257, 258 cycles 261 mass 259–261 protein synthesis 261–266 shutdown 261 food changes in composition 12–14 choice of 11–14 adaption 4 timescales 6–7 flavour 4–5 sensation and metabolism 3–8 single 11–12 food aversion 4, 5 food intake conditioning stimuli 8–11 cyclical 11–12 daily 13–14, 13 modelling 21, 22 learned associations 3–8 mathematical models 21–24, 35–36 prediction of intake 24–31

metabolic control 24 photoperiodic effects 205–206 voluntary 22, 24 forage 7, 12, 30 forebrain 54 fractional outflow rate 28 frontal cortex 45 frugivores 62 fungal glycoside hydrolases 83–84, 86–87 fungi 64, 79, 86, 87

gastric motility control sites 41–42 brainstem 47–48 brain stimulation 48 interior 50–52 mechanism 54–56, 55 single unit recording 48–49 surface sites 49–50, 49, 53 forebrain 54 gastrointestinal metabolism 131–144 gastrointestinal parasites 425 defence mechanisms 426–429 effect on host 429–430 gastrointestinal tract anatomy and vasculature 131–132 energy metabolism 134–135 measurement of blood flow 132–133 microbial community 64–65 urea 156–157, 156 use of labelled substrates 133–134 gazelles 172 GDH (glutamate dehydrogenases) activity 109 gene transfer 87, 390–394 barriers to 393–394 evolution of rumen microorganisms 394 genetic effects hyperplasia 240 hypertrophy 246–247 wool growth 260–261 genetic fingerprinting 69 genetic manipulation 373–375 digestion 382 disease resistance 382–383 endocrine system 376–378 intermediary metabolism 378–382 genetic selection 413–415, 414 genomic analysis 399 genotype–nutrition interactions 412–413

Index

467 glucanases 86 glucocorticoids 454 glucose blocking uptake 9 metabolism absorption 139–140 availability 140–141 gut tissues 137–138, 138 release 139–140 glucose co-transporter gene 140 glucose irreversible loss (GIL) 138, 138 glutamate dehydrogenase (GDH) 152 glutamine 143, 153, 161, 161–162, 455–456 Gly-Arg-MNAse 105–106 glycanase genes 87 glycoside hydrolases 83–84, 86–87 glycosyl hydrolases 88–89 goats 4, 7, 48 grazers 62 Great Lakes, North America 63 growth hormone circulating 188, 195 effect on metabolism 196–198 galactopoeitic effects 216 GH somatotropin 187 hyperplasia 238 receptors (GHR) 188–189 growth hormone binding protein (GHBP) 189 growth regulation see postnatal development gut ecology DNA-based studies 63–64, 65 electrophoresis DGGE 69–70 TGGE 69–70 RNA-based studies 65–67 gut tissues amino acid absorption 143 amino acid requirements 141–143

hamsters 207, 211 heart 176–177 hepatic GHR 194 hepatic protein turnover 162–164 hepatic urea 156 hepatocytes 150, 151, 153 herbivores 62–64 hindgut 25 homeorhetic regulation 286–288 homeothermy 171–172

hormones, photoperiodic effects 211–217 host resistance blowfly strike see blowfly strike fleece rot 437 gastrointestinal parasites 425, 429–430 defence mechanisms 426–429 protective immunity 430–431 mastitis see mastitis humans 4, 47, 120–123 bacterial diversity 70, 71 colonic ecosystem 68 humoral responses (resistance) 427–428 hunger 10 hydrogen transfer 100 hydrolase synthesis 89 hypermotility 53–54 hyperplasia 238–240, 239 hyperthermia fetal 299–300 maternal 301–303 hypertrophy 233–234 control and manipulation exogenous hormones 243–246 genetic effects 246–247 postnatal nutrition 246 hypoglycaemia 197–198 hypothalamus 10, 54

IGF binding proteins 190–192 receptors 192 IGF (insulin-like growth factor) 329, 330 IGFBP proteases 333 IGFBP-3 332 IGFBPs 330–332, 331 lactoferrin 342–343 mammary physiology blood versus milk changes 333–336, 334, 336 IGF receptors 337–339 IGFBP 196, 339–340, 340 IGFBP-3 binding proteins 340, 342, 342 and IGF-I 341 and illness 198–199 nuclear appearance 343–344 nuclear localization sequence (NLS) 343 local IGF ligands 337 receptors 329–330

Index

468 IGF-I 164, 187, 196–198 endocrine 189 gene 190 plasma levels 194 IGF-II 194 illness 198–199 immune system adipose tissue interactions 177–178 genetic influences 453–454 influence of nutrition 455–457 influence of stress 454–455 leptin 180–181 immunity non-specific enhancement 431–432 protective 430–431 immunological functions 453–454 immunosuppression 454–455 inductive phenomena 426 inflammatory mediators 428 inflammatory process 451, 452–453 ingestion 6–7 insulin 164, 175, 177, 198, 319 insulin-like growth factors 189–190 interferons (IFN) 452 interleukin-6 173 interleukins (IL) 452 interspecific differences 27 intramuscular adipocytes 177 intraruminal infusion 4 intraruminal nitrogen recycling 102–104 isotopic labelling 134 isotopomer analysis 152

keratin 450 kinetics, digestion 24–31

Lachnospira multiparus 66 lactation 264 adaptations to 314–317, 315, 318 mechanisms 317–320 dairy cattle see dairy cattle nutrient partitioning 311 coordination 320–324 endocrine control 415–418 regulation 312–314 photoperiodic effects 215–217 Lactobacillus 71 lactoferrin 342–343 lambing 298, 299

lambs 4, 49 dietary choice 5–6, 7–8, 11 gastric motility 49 larkspur 11 learned associations food flavours 4–5 food intake 12–14 food and metabolism 3–8 leptin 9, 10, 174, 175, 178–181, 182 Leucaena leucocephala 66 leucine 143 leucocytes 449 distribution 450–451 function 451 proportions 453, 454, 456 LGMs (large ground mammals) 62 LiCl aversion studies 4, 7 lignin 91–92 limbic system 10 lipase 207 lipogenesis 357 lipolysis 175–176, 178, 196–197 milkfat output 348, 357–358 lipoprotein lipase 174 liver amino acid extraction 159, 160 amino acid metabolism see amino acids ammonia metabolism see ammonia metabolism architecture 149–151, 151 fatty acids 177 hepatic protein turnover 162–164, 162, 163 link to CNS 8 substrate supply 131–144 ureagenesis see ureagenesis lumen 139, 429 lymph nodes 177–178 lymphocytes 450–451 lysing 177

macronutrient partitioning 275–288 macrophages 450, 451 mammalian herbivores 62–64 mammary cells adipose tissue 176 defence mechanisms 449–453 IGFBP-3 binding protein see IGF peptide utilization 126 mammary gland in vivo 127

Index

469 tissue explants 126–127 photoperiodic changes 215–217 Mammut americanum 63 mandibular secretions 42–44, 43 mastitis genetic influences 453–454 host resistance 449–450 inflammatory process 452–453 leucocyte functions 451 leucocyte populations 450–451 maternal tissues 284–288 mathematical models 21–23 dairy cattle metabolism 363–366 digesta kinetic 24–31 inadequate detail 23 intake and digestion 24–31, 35–36 metabolic 31–36 ME (metabolizable energy) 6, 14, 33 meal patterning 34 meat quality 247–248 mechanistic models 22–23 medulla oblongata 9 melatonin 211–212, 213 mesenteric drained viscera (MDV) 138 mesodermal cells 227–229 metabolic constraints 23–24, 34–35 ‘metabolic discomfort’ 9–10 metabolic imbalance 5–6, 198–200 metabolic models 21 animal response to nutrients 33–34 constraints on intake 34–35 microbial metabolism 31–33 metabolic pathways 33 metabolic signals 3 metabolic stress 455–456 metabolic zonation 151, 151 metabolism amino acid see amino acids ammonia see ammonia metabolism energy 196–198 gastrointestinal tract 134–135 genetic manipulation 378–382 glucose 137–141 and learning see learned associations microbial 31–33 muscle 206–208, 208 nitrogen see nitrogen postnatal development see postnatal development protein 162, 196–198 short-chain volatile fatty acids 135–137

methanogenic bacteria 66–67 mice 71 Michaelis–Menten kinetics 31–32, 33 microbes endosymbiotic 63 interactions with plant cells 79 outflow rates 32 rumen 6, 7 microbial digestion 62–63 microbial diversity 67–69 microbial ecology 100–104 microbial metabolism 31–33 microbial physiology 104–108 micronutrients 456–457 milk fat content 324, 356 leucocyte populations 450–451 secretion, sheep 208–211, 209, 210, 211 Miocene period 62 mitochondria 150, 151 molecular adhesion 80, 86 molecular ecology 61–72 monensin 101 monosaccharide analysis 83 MP (metabolizable protein) 33 mRNA, tissue distribution 119–120, 194 mucosal bacteria 71 mucus 429 mule deer 8 multi-catheterization procedures 132 multi-enzyme complexes 81, 86 muscle fibre number 240 muscle growth determination 227–229, 228, 229 differentiation 232–233, 238 hyperplasia see hyperplasia hypertrophy see hypertrophy myogenic cells 237 proliferation 229–232 skeletal 237 muscle metabolism 206–208 mutational analysis 104–107 myoelectric activity 28 myogenic cells 125–126

NACl infusion 44–46 nematodes 426–429 Neocallimastix 67, 82 net energy (NE) 31

Index

470 neuropeptides 428 neutral detergent fibre (NDF) 15, 28, 135 neutrophils 450, 451, 452, 455 New Zealand 101 nitrogen control 108–110 nitrogen metabolism liver see liver microbial ecology 99–104 microbial physiology genetic examination 104–108 nitrogen recycling 102–104 non-adherent mutants 85–86 non-shivering thermogenesis (NST) 306 NTS (nucleus of the solitary tract) 9 nuclear localization sequence (NLS) 343 nucleases 393 nucleic acid probes 64–65 nutrient flux, glucose availability 132, 140–141 nutrient partitioning lactation see lactation pregnancy see pregnancy sheep 208–211, 209, 210, 211 nutrient ratio 5–6 nutrient requirements 10–11, 12–14 nutrient–gene interactions 409–410, 418–419 genotype–nutrition interactions 412–413 prediction of growth responses 411–412, 412 nutrients animal response, models of 33–34 choice of food 4, 5–6, 10–11 control of hyperplasia 238 deficiency 4–5 essential 4–5 excess 5, 9–10 flow rates 28 receptors 9–10 sufficiency 5 nutrition and immunity 455–457 postnatal 246 prenatal 241–243 somatotropic axis 195–196 wool growth see wool growth ‘nutritional wisdom’ 5

oaten chaff 13

obesity 175 obligate anaerobes 63 obligate peptide bacteria 100–101 oestrogens 173–174 oligonucleotide probes 71, 107 omasum 140 operational taxonomic units (OTUs) 68 opsonins 451 ornithine cycle 152–154, 153 substrate priority 154–155 osmosensitive system 44–47 Ostertagia 8 ovine peptide transporters functional characteristics 117–119 structure 120 tissue distribution of mRNA 119–120 OVLT (organum vasculosum, lateral terminalis) 45, 46 oxygen consumption 134–135

Palocene period 62 paracrine effects 174–177 parasitism 8 parotid nerve 46 parotid secretions 42–44, 43 inhibiting 44–47, 46 stimulating 45 particle kinetics 28 particulate outflow 27–28 particulate pools 28 parturition 14, 454 passage kinetics 27 rate 30–31 PCR 101–102 PCR-amplified DNA 69 peptidases 105 peptides absorption 123–125 transport 117–123 utilization 125–127 Peptostreptococcus 101 Peptostreptococcus anaerobius 100 perinatal transition 193 periportal cells 150 periportal hepatocytes 152 peristalsis 429 perivenous cells 150 PFGE (pulsed-field gel electrophoresis) 103

Index

471 pH peptide transport 118–119 ruminal fluid 136 phagocytosis 451 phenolics 7–8 phenomenological models 22–23 phlorizin 140 photoperiodic effects 218 adipose tissue 206–208, 208 energy expenditure 206 food intake 205–206 hormonal mediation adipose tissue 213–215 brain 211–212 mammary gland 215–217 peripheral hormones 212–213 milk secretion 208–210, 209, 210, 211 nutrient partitioning 210–211, 212 sheep 205–219, 218 phylogenetic analyses 101 pigs 70 pituitary hormones 180 placental lactogen (PL) 188 placental substrate 193 plant cell walls microbial degradation 89–92, 91 microbial interactions 79–86 plasma amino acids 157–158 catecholamines 206–207 GH concentrations 193, 195 IGF-I levels 194, 195 IGFBP-1 194–195 insulin 198 plasma export proteins 160 plasmids 392–393, 392 polyclonal antibodies 85 polyethylene fibre 7 polyethylene glycol (PEG) 8 Polyplastron multivesiculatum 67 population analysis 101 porta hepatis plexus 131–132 portal-drained viscers (PDV) 138 portal flux peptides 123–125 SCVFA 136–137, 137 portal vein blood flow 134 positive feedback 34 post-ingestive feedback 6 postnatal development 187–188 growth regulation

somatotropic axis circulating growth hormone 188 developmental changes 192–193, 194–195 growth hormone receptors 188–189 IGF binding proteins 190–192 IGF receptors 192 influence of nutrition 195–196 insulin-like growth factors 189–190 role in metabolic imbalance 198–200 metabolism effects of GH and IGF-I 196–198 postprandial hypertonicity 47 predominant ruminal bacteria (PRB) 101 pregnancy 264 macronutrient partitioning amino acids 276–277 fetal nutrient requirements 275–276 glucose 276–277 maternal nutrition 279, 282–283 direct effects 277–278 indirect effects 278–279 maternal–fetal transfer 281–282 non-fetal conceptus tissues 276 maternal tissues homeorhetic regulation 286–288 metabolic adaptations amino acid metabolism 284 glucose metabolism 284 modulation by nutrition energy 284–285 protein 285 placenta metabolism amino-acid 281–282 glucose 281, 283 nutrient transport amino acids 280–281 glucose 280 size 277–279, 282 preterm labour 308 Prevotella bryantii 65, 104–107 Prevotella ruminicola 68 Prevotella/Bacteroides 397–398 probabilistic models 28 prolactin (PRL) 188

Index

472 propionate 5, 8, 136, 153, 155 prostaglandin E2 175 protein binding 174, 190–192 crude (CP) 14, 28 metabolism 162, 196–198 requirements 10–11 supply 6 synthesis hepatic 163 wool see wool growth proteolytic bacteria 101–102 proteolytic enzymes 108 protozoa 32, 66–67, 79, 102 protozoal lysis 102–103 protozoan glycoside hydrolases 83–84, 86–87 PrtA activity 106

genetics 389 exploitation of genes 399–400, 400 gene expression 140, 394–400 gene transfer 390–394 genomic analysis 399 regulation 398–399 vector systems 394–398, 395 models of 23, 25 rumen inoculants 401 rumen microbial diversity 67–69 rumen osmolality 7 rumination 53–54 Ruminococcus albus 66, 69–70, 79, 82, 107 Ruminococcus flavefaciens 66, 69–70, 79, 82 Ruminococcus spp. 397

rabbits 120–123 ratio of nutrients 5–6 rats 4, 47, 120–123 rDNA sequencing 63 RDP (rumen degradable protein) 6 receptors, nutrients 9–10 regression model 22 reindeer 172, 206 reiterated sequences 82 resistance see host resistance restriction modification systems 393 retention timescales 31 reticular contractions 49–50, 49, 51–52, 52, 53 reticular/oesophageal groove 52 reticulo-omasal orifice 52 reticuloruminal motility 54–56, 55 retinol binding protein 174 RNA-based studies, gut ecology 65–67 rumen function contractions 28, 49–51, 50, 51 fermentation 5–6 fungi 64, 66–67, 86, 87 maintenance of 7 microorganisms adherence to plant cell walls 79–86 cellulose-binding proteins 82, 85 cellulosome equivalent 81–82 non-adherent mutants 83, 86 diversity 64–65, 67–69 effect of diet 70

16S rDNA 63, 70 16S rRNA 64, 65–66, 68, 69, 71, 100 18S rRNA 67 salivation control sites 41–47, 43 amygdala, osmosensitivity 44–47 brainstem 42–44, 43 experiments 41–42 frontal lobe 44, 45 satiety 10, 12 scaffoldins 80, 81 scaling (models) 27 Scotland 172, 207 selection programmes 453, 454 selenium (Se) 456 Selenomonas ruminantium 65, 394 septic shock 199 sequence-based analysis 67–69 SGLT1 Na+ 140 SGMs (small ground mammals) 62 sheep 13 blowfly strike see blowfly strike brainstem 50–54, 50, 51, 53 change of diet 13–14 ewes 10 food preferences 4, 5, 6, 7 gastric motility 41–42, 47–56 gastrointestinal parasites 425–433 GH secretion 194–195 immunity 440–441 liver 154 mastitis see mastitis

Index

473 parotid secretion 45, 46 photoperiodic effects see photoperiodic effects portal vein 132 salivation 41–47, 43 septic shock 199–200 wool growth see wool growth short-chain volatile fatty acids 135 absorption 136 metabolism by gut tissues 136, 137 signature probes 67 single unit recording 48–49 small intestine 25 sodium bicarbonate 7 somatic cell count (SCC) 453, 456 somatostatin (SRIF) 188 somatotropic axis 187, 198–200 developmental changes 192–193 perinatal transition 193 postnatal ontogeny 194–195 influence of nutrition 195–196 metabolic imbalance 198–200 regulation of postnatal growth 188–192 splanchnic bed 131–134, 158 Staphylococcus aureus 452 starch 4–5 starch digestion 139–140, 139 statistical models 22–23 Streptococcus 71, 102 Streptococcus bovis 65, 398 stress 454–415 submandibular gland 45 submandibular secretions 42–44, 43 substrate fractionation (food) 25–26 substrate supply, liver 131–144 substrates, labelled 133–134 sulphur 4 surface erosion (cell walls) 90–92, 91 Synergistes jonesii 66

T-cells 450, 451, 455 TA-cloning 107 taste afferents 10 teats 450 temperature, body see body temperature tetracycline resistance 390 TGGE (temperature gradient gel electrophoresis) 69–70 theory of 3–17 thermal physiology, fetus see fetus

thermoregulatory mechanisms 305–307, 306 toxicosis 6–7 toxins 4, 7–8, 11, 12, 62 carbon metabolism 108, 109 tracer techniques 141–143 Trangie flock 438–440 transgenic technology 374–375, 377 transonic flow probes 133 transposable elements 393 transposon mutagenesis 104 triacylgycerol 171 trypticase 109 tumour necrosis factor (TNF) 173, 175, 452

udder diseases 449–458 udder infections see mastitis undernutrition 196 urea production 153, 161 ureagenesis 154, 155–157 control of 155 synthsized urea 156–157, 156

vaccines 441–444 venous–arterial (V–A) differences 132–134 villus 143 visceral afferents 10 visceral organs 8–9 vitamin A 457 vitamin E 456–457 VLDLs (very low density lipoproteins) 174 volatile fatty acids (VFA) 32

white blood cells 455 whole-cell hybridization 71 wool growth models follicle 257–258, 258 nutritional 258–259 protein synthesis 259 protein synthesis 255–257, 256 skin and follicles control of mass 259–260 genetics 260–261 hormonal 261 nutrition 260 protein synthesis 259, 261–262

Index

474 wool growth continued skin and follicles continued protein synthesis continued amino acids dietary supply 262–263, 263 growth 264 hormonal effects 264–265 pregnancy and lactation 264

nutrient supply 266 proportion in wool 266–267, 266 xylanases 86 xylosidase activity 109 yellow affinity substance (YAS) 89 zinc (Zn) 457